! i!

Long-Term Risk of New-Onset Diabetes in Pediatric Solid Organ Transplant Recipients

by

Dr. Rahul Chanchlani

A thesis submitted in conformity with the requirements for the degree of Master of Science (Clinical Epidemiology)

Institute of Health Policy, Management and Evaluation University of Toronto

© Copyright by Rahul Chanchlani 2016

! ii!

Long-Term Risk of New-Onset Diabetes in Pediatric Solid Organ Transplant Recipients

Rahul Chanchlani

Master of Science (Clinical Epidemiology)

Institute of Health Policy, Management and Evaluation

University of Toronto

2016

ABSTRACT Background: Precise estimates of diabetes mellitus in transplanted children are not known.

Objective: Determine the risk of diabetes in pediatric transplant recipients and compare the risk

with non-transplanted children. Methods: Transplanted (n=1020) and non-transplanted children

(7, 134,067) children were linked with provincial health administrative data at ICES to determine

rates of diabetes. Results: Overall, the incidence rate ratio of diabetes was 7-times higher in the

transplant (IRR 7.0, 95% CI:5.9, 8.3) than the non-transplant cohort. The transplant cohort had a

9-fold (HR 8.9, 95% CI:7.5, 10.5) higher hazard of diabetes than the non-transplant cohort. Risk

was highest within the first year of transplant but remained elevated even after 5-10 years. Lung

and multiple organ groups had a 5-fold (HR 5.4, 95% CI:3.0, 9.8) higher hazard compared to

kidney transplant recipients.

Conclusions: Risk of diabetes is higher in transplanted children with lung and multiple organ

transplant recipients at the highest risk.

! iii!

ACKNOWLEDGEMENTS

I am very thankful for the guidance and assistance provided by my thesis committee members:

Dr. Rulan Parekh, my supervisor, for always being available, for providing great mentorship, and

constant inspiration, Dr. Joseph Kim for his content and methodological advice, and Dr. Vanita

Jassal for her thoughtful and constructive feedback.

The Hospital for Sick Children’s Division of Nephrology for providing excellent nephrology

training.

My research team members, Jovanka, Tonny, Karlota, and Esther without whom this work

would not have been possible.

I would also like to thank the funding sources I received for the Masters program,

including the Restracomp Research award from the Research Training Center at the

Research Institute, Hospital for Sick Children, Toronto and a Transplant Fellowship from the

Transplant and Regenerative Medicine Center at the Hospital for Sick Children, Toronto.

I also thank the Institute for Clinical and Evaluative Sciences, Toronto

for providing me access to the data for this study, and

Dr. Stephanie Dixon who helped me immensely through this project.

I thank my grandparents, parents, wife (Dr. Ritika Arora) and my lovely daughter, Aishi for

providing affection, inspiration and support over the many years.

! iv!

TABLE OF CONTENTS

Abstract………………………………………………………………………………….ii Acknowledgements……………………………………………………………………..iii Table of Contents……………………………………………………………………….iv List of Tables……………………………………………………………………………vi List of Figures…………………………………………………………………………..viii 1 Introduction…………………………………………………………………………..1

2 Background

2.1 Burden of diabetes in Canada in the general population……………………………2

2.2 The health impact of diabetes……………………………………….........................5

2.3 The economic impact of diabetes…………………………………………………...5

2.4 Burden of diabetes in solid organ transplant recipients……………………………..5

2.5 Pathophysiology and risk factors of diabetes in solid organ transplant recipients….6

2.6 Diagnosis of diabetes mellitus………………………………………………………13

2.7 Use of health administrative data to diagnose diabetes……………………………..14

2.8 Sequelae after developing diabetes in solid organ transplant recipients……………15

3 Rationale……………………………………………………………………………...17

4 Objectives and Hypotheses………………………………………………………….18

5 Methods

5.1 Setting and participants……………………………………………………………..19

5.2 Inclusion criteria…………………………………………………………………….19

5.3 Exclusion criteria……………………………………………………………………20

5.4 Data sources and linkage……………………………………………………………20

5.5 Outcome assessment and classification……………………………………………..21

5.6 Study timeline……………………………………………………………………….23

5.7 Study cohort assembly………………………………………………………………23

5.8 Defining covariates………………………………………………………………….24

5.9 Statistical and sensitivity analysis…………………………………………………...25

5.10 Power calculation…………………………………………………………………..27

5.11 Ethical consideration……………………………………………………………….28

! v!

6 Results

6.1!Study cohort……………………………………………………………………………….29

6.2!Baseline characteristics of the exposed and unexposed cohort……………………….30

6.3 Baseline characteristics of the individual pediatric organ transplant groups……….31

6.4 Incidence rate of diabetes in the exposed and unexposed cohort…………………...32

6.5 Incidence rate of diabetes in the exposed and unexposed cohort, stratified by era…34

6.6 Incidence rate of diabetes in the individual pediatric organ groups………………...35

6.7 Relative risk of diabetes in exposed and unexposed cohort………………………...36

6.8 Relative risk of diabetes in exposed and unexposed cohort, stratified by era………38

6.9 Risk of diabetes in the individual organ groups…………………………………….41

6.10 Transplants in exposed and unexposed groups after the index date……………….44

6.11 Sensitivity analysis………………………………………………………………...45

7 Discussion 7.1 Summary of findings………………………………………………………………..49

7.2 Strengths…………………………………………………………………………….53

7.3 Limitations…………………………………………………………………………..53

7.4 Implications…………………………………………………………………………54

7.5 Conclusions…………………………………………………………………………55

7.6 Future Directions……………………………………………………………………56

8 Acknowledgements for using ICES data…………………………………………..57

9 References……………………………………………………………………………58

Appendix 1 RECORD checklist……………………………………………………....63

! vi!

List of Tables

Table 1 American Diabetes Association criteria for the diagnosis of impaired glucose

tolerance and diabetes

Table 2: Studies on use of health administrative data for identifying diabetes in children

Table 3: Exclusion of individuals in transplant (exposed) and non-transplant (unexposed)

cohort

Table 4: Baseline characteristics of the transplant and non-transplant cohort

Table 5: Baseline characteristics of the individual pediatric solid organ transplant groups

Table 6: Incidence rate of diabetes in exposed and unexposed cohort overall and at specific

time intervals

Table 7: Incidence rate of diabetes in exposed and unexposed cohort overall and at specific

time intervals, stratified by era

Table 8: Incidence rate of diabetes in individual organ groups at specific time intervals

Table 9: Hazard ratio for diabetes in exposed and unexposed cohort overall and at specific

time intervals

Table 10: Hazard ratio for diabetes in exposed and unexposed cohort overall and at specific

time intervals, stratified by era

Table 11: Hazard ratio for diabetes in individual organ groups overall and at specific time

intervals

Table 12: Risk of diabetes exposed and unexposed cohort after excluding lung and multiple

organ transplant groups

Table 13: Risk of diabetes in exposed and unexposed cohort after censoring individuals

receiving a transplant after the index date

Table 14: Risk of diabetes in individual organ groups after censoring individuals receiving

! vii!

a transplant after index date

! viii!

List of Figures

Figure 1: Age-standardized prevalence and number of cases of diagnosed diabetes among

individuals aged one year and older, Canada, 1998/99 to 2008/09

Figure 2: Prevalence of diagnosed diabetes among individuals aged one year and older, by

age group, Canada, 1998/99 to 2008/09

Figure 3: Pathophysiology of new-onset diabetes after transplant Figure 4: Risk factors associated with development of new-onset diabetes post transplant Figure 5: Linkage of pediatric transplant cohort with ICES datasets Figure 6: Study timeline Figure 7: Log-rank test analysis showing the detectable hazard ratios for diabetes in

transplant cohort compared to the unexposed cohort

Figure 8: Study cohort after meeting the inclusion and exclusion criteria

Figure 9: Kaplan Meier curve for risk of diabetes in transplant and non-transplant cohort

Figure 10: Cumulative risk of diabetes in transplant and non-transplant cohort in era

1(1991-2002, a) and era 2, (1993-2014, b)

Figure 11: Cumulative risk of diabetes in individual pediatric organ transplant recipients

Figure 12: Cumulative risk of diabetes in exposed and unexposed cohort after excluding

lung and multiple organ recipients

! 1!

1.! INTRODUCTION

Diabetes mellitus (i.e., “diabetes”) is an important public health issue due to its increasing

burden, associated economic costs, and long-term complications (1). Based on the public health

agency of Canada 2011 annual report (2), almost 2.4 million Canadians (6.8%) were living with

diabetes in 2008 to 2009. Over the last two decades, the incidence of diabetes has increased not

only in adults but also in children and adolescents, primarily due to changes in the lifestyle,

dietary habits and increase in the prevalence of obesity.

In addition to the general population, diabetes is also an increasingly recognized

complication among solid organ transplant recipients in adults. Organ transplantation is a

lifesaving therapy for many types of organ failure, including kidney, liver, heart, lung, and small

bowel (3). Although not curative, transplantation leads to significant improvements in quality of

life and also increases survival in children. Advances in immunosuppression have significantly

improved graft and patient survival after solid organ transplantation. Improved survival,

however, predisposes children to multiple comorbidities after transplantation from long-term

exposure to the underlying cause of end-organ failure, transplant medications, and the likely

need for re-transplantation over a lifetime. Among the various complications experienced by

organ transplant recipients, new-onset diabetes after transplantation is a major clinical problem

that affects long-term graft and patient survival due to cardiovascular and infectious

complications (4, 5).

It is imperative to study the burden of diabetes and its associated risk factors in pediatric

solid organ transplant recipients to allow for timely diagnosis, appropriate health resource

allocation, and improve graft and patient survival.

! 2!

2.! BACKGROUND

The following sections are a discussion of relevant background describing the burden and

impact of diabetes in healthy individuals as well as those with a solid organ transplant. It also

highlights what has been studied previously, elaborates the gaps in knowledge, and provides the

rationale for the current study.

2.1 Burden of diabetes in Canada in the general population

Based on the public health agency of Canada 2011 annual report (2), the prevalence of

diabetes has increased significantly in the last decade. In fact, from 1998 to 2009, the prevalence

of diagnosed diabetes among adult Canadians rose by 70% (Figure 1). The prevalence of

diabetes in the 35 to 39 and 40 to 44-year age groups doubled, primarily due to increasing rates

of obesity (Figure 2). It is estimated that approximately 4 million Canadians will suffer from

diabetes by 2019 (2).

Diabetes has also become a common chronic disease among children and youth.

Recently, type 2 diabetes, which is typically considered as an adult disease, has been on the rise

throughout the world in children and adolescents for two decades. Moreover, the burden of type

1 diabetes among children between 1 to 9 years of age has also increased, from 0.1% in 1998/99

to 0.2% in 2008/09 (Figure 2) (2).

! 3!

Figure 1: Age-standardized prevalence and number of cases of diagnosed diabetes among individuals aged one year and older,

Canada, 1998/99 to 2008/09

Source: ©All Rights Reserved. Diabetes in Canada: Facts and figures from a public health perspective. Public Health Agency of

Canada, 2011. Reproduced with permission from the Minister of Health, 2016

! 4!

Figure 2: Prevalence of diagnosed diabetes among individuals aged one year and older, by age group, Canada, 1998/99 to

2008/09

Source: ©All Rights Reserved. Diabetes in Canada: Facts and figures from a public health perspective. Public Health Agency of

Canada, 2011. Reproduced with permission from the Minister of Health, 2016

! 5!

2.2 The health impact of diabetes

Diabetes has a significant health impact and can lead to many short- and long-term

complications, including cardiovascular disease, hypertension, vision loss, chronic kidney

disease, and neuropathy. Persons with diabetes are 3 times more likely to be hospitalized with

cardiovascular disease, 12 times more likely to be hospitalized with end-stage renal disease and

twice as likely to die than those without diabetes (2). In fact, diabetes is responsible for one-third

of incident cases of end-stage renal disease, thereby creating an increasing demand for renal

replacement therapy and kidney transplantation (2). It is estimated that more than one in ten

deaths in Canadian adults could be prevented if diabetes rates were reduced to zero (2).

2.3 The economic impact of diabetes

Individuals with diabetes are three times more likely to have been hospitalized at least

once and have a longer hospital stay than those without diabetes (2). Data from the Public Health

Agency of Canada's Economic Burden of Illness in Canada (EBIC) report, provided a

conservative estimate of $2.5 billion in the year 2000 for the total cost of diabetes, excluding cost

associated with diabetes-related complications (2). Moreover, it is expected that costs will only

continue to rise with the increasing prevalence of diabetes.

2.4 Burden of diabetes in solid organ transplant recipients

New onset diabetes mellitus after transplantation is a common complication in both

pediatric and adult solid organ transplant recipients. (6). Diabetes occurs most frequently in the

first year after transplantation when patients are treated with the highest doses of

immunosuppression to prevent rejection. Among adults, diabetes has been reported in 4 to 25%

kidney transplant recipients, 2.5 to 25% liver transplant recipients, 4 to 40% heart transplant

! 6!

recipients, and 30 to 35% lung transplant recipients (7, 8). Similarly, there is also quite a

significant variation in the incidence of diabetes in children with solid organ transplants. It varies

from 3 to 20% in pediatric kidney transplant recipients (9, 10), 8 to14% in liver transplant (11,

12), 4 to 40% (13) in heart and heart/lung transplant recipients. The variation in the reported

incidence is most likely due to the lack of a standard definition across studies, the duration of

follow-up, the presence of both modifiable and non-modifiable risks factors, and the specific

organ transplanted (14).

2.5 Pathophysiology and risk factors of new-onset diabetes in solid organ transplant

recipients

The central pathophysiological defect occurring in new onset diabetes after transplant is a

failure of pancreatic β cells to compensate for insulin resistance, which is similar to type 2

diabetes mellitus (Figure 3) (15, 16). There are, however, a number of modifiable and non-

modifiable risk factors that also influence the development of diabetes post-transplant (Figure

4). It is important to understand these additional risk factors so that appropriate clinical decisions

can be made promptly.

! 7!

Figure 3: Pathophysiology of new-onset diabetes after transplantation

Insulin'resistance,'decreased'insulin'

secretion,'and'beta'cell'damage'

Decreased'glucose'uptake

Increase'lipolysis

Increase'glucose'production

Hyperglycemia

Diabetes'mellitus

Modifiable'risk'factors

Non;modifiable'risk'factors

! 8!

Figure 4: Risk factors associated with the development of new-onset diabetes post transplant

2.5.1 Non-modifiable risk factors associated with post-transplant diabetes

a) Age: In adult transplant recipients, advancing age has been shown to be associated

with the development of new-onset diabetes post-transplant, similar to that in the general

population. A study on 11,650 kidney transplant recipients using USRDS data (United States

Renal Data System) showed that the risk of diabetes increased with age and patients > 60 years

were at the highest risk of diabetes (17). However, a specific age cut-off beyond which there is

an accelerated risk of diabetes has not been established (16).

Post%TransplantDiabetes0

CVD,0Infection,0and0Mortality

Modifiable0Risk0Factors% Obesity% Immunosuppressive0medications% Hypomagnesemia0and0Hypokalemia

% Socio%economic0status% Growth0hormone

Non%Modifiable0Risk0Factors% Age% Ethnicity% Primary0end0organ0disease% Family0history0of0diabetes% Genetic0susceptibility0

Pediatric0Solid0Organ0Transplant0Recipients

! 9!

b) Ethnicity: Similar to adults, children of African-American ethnicity are at an

increased risk of diabetes after thoracic transplantation (18). Data from the North American

Pediatric Renal Transplant Cooperative Study (NAPRTCS) suggested that African-Americans

were at a higher risk and Hispanics were at a lower risk of developing diabetes as compared to

Caucasians (9).

c) Primary organ disease: Some children are predisposed to diabetes by virtue of their

primary disease leading to organ failure. For example, in a study of 77 lung transplant recipients,

43 had cystic fibrosis. Before lung transplantation, diabetes was diagnosed in 63% of individuals

with cystic fibrosis as compared to 6% without cystic fibrosis. After transplantation, 60% of

patients with cystic fibrosis developed new-onset diabetes after transplant (19). Similarly,

hemolytic uremic syndrome (20) and autosomal dominant polycystic kidney disease (21) also

predispose individuals to diabetes due to pancreatic insufficiency from the underlying disease

based on activity leading to insulin deficiency and insulin resistance, respectively.

d) Family history of diabetes: In the general population, a family history of diabetes is

significantly associated with an increased risk of diabetes. However, the evidence is conflicting

in individuals with a solid organ transplant. A few studies have shown upwards of a 7-fold

increased risk of diabetes secondary to a family history (22), but others have not shown any

association with the risk of post-transplant diabetes (23).

e) Genetic susceptibility: There are a number of genetic polymorphisms detected in the

development of new-onset diabetes after transplant. For example, Transcription factor 7-like 2

(TCF7L2) factor has been found to be different in those with and without diabetes (24). The

polymorphism in the TCF7L2 leads to decrease in the insulin secretion and increased glucose

production in liver. It has been shown to be associated with new-onset diabetes within 6 months

! 10!

of transplant in a large cohort Caucasian (25) and Korean (26) kidney allograft recipients.

Similarly, IL-6 promoter polymorphisms have been shown to affect the risk of diabetes after

transplant. An increase in the IL-6 secretion is associated with a decreased risk of diabetes (27).

The implications of genetic risk factors suggest that screening at the time of transplant may be

warranted in the future.

2.5.2 Modifiable risk factors associated with development of diabetes post-transplant

a) Immunosuppressive medications: Tacrolimus, corticosteroids, cyclosporine,

sirolimus, and various induction therapies are commonly used across all solid organ groups and

are well known to have diabetogenic effects.

Corticosteroids have a dose-dependent effect on diabetes. An increase in dose of

prednisone of 0.01 mg/kg/day is associated with a 5% increase in the risk of diabetes after

transplant (22). Glucocorticoids increase glucose production in the liver, reduce peripheral

glucose uptake, and decrease insulin sensitivity (28). A recent meta-analysis showed significant

improvements in hyperglycemia and cardiovascular risk factors in adult kidney transplant

recipients who did not receive corticosteroid treatment (29). In contrast, 130 children with a

kidney transplant did not have different rates of diabetes with a steroid-based compared to

steroid-free immunotherapy in a randomized controlled trial (30).

Calcineurin inhibitors (CNI) such as tacrolimus and cyclosporine cause diabetes after

transplant by a number of mechanisms, including reduced insulin secretion, increased insulin

resistance, or through a direct islet cell damage (31). Similar to corticosteroids, the diabetogenic

effect of calcineurin inhibitors is dose-dependent (32). Compared to cyclosporine, tacrolimus is

more diabetogenic. In children with kidney transplants, tacrolimus-based therapy led to diabetes

! 11!

in 35% compared to 9% in cyclosporine-based treatment. On switching from tacrolimus to

cyclosporine within six months, diabetes resolved in 50% of cases (10). The DIRECT (Diabetes

Incidence after Renal Transplantation) study, a randomized controlled trial, also confirmed the

greater diabetogenic effect of tacrolimus. Approximately, 700 kidney transplant recipients were

randomized to receive either tacrolimus or cyclosporine. At 6 months post-transplant, the

incidence of diabetes was higher in the former group (33.6% versus 26.0%; P = 0.046) (33). A

recent study investigated the potential reason for higher diabetogenic effect of tacrolimus than

cyclosporine in Zucker rats and concluded that tacrolimus leads to a greater inhibition of Ins2

gene and beta cell proliferation in special environment of insulin resistance (34). This effect was

not observed in insulin-sensitive rats. These symptoms recovered after a short withdrawal of

CNI’s.

Treatment with the mammalian target of rapamycin inhibitor, sirolimus, has also been

identified as a risk factor for the development of diabetes post-transplant (35). Teutonico et al.

demonstrated that the withdrawal of calcineurin inhibition and the conversion to sirolimus

resulted in a worsening of insulin resistance (36). Sirolimus-induced hyperglycemia is attributed

to the inhibition of the pancreatic β cell response and increased peripheral insulin resistance (37).

Basiliximab, an IL-2 receptor antagonist, leads to depletion of regulatory T-lymphocytes

that helps in maintaining immunologic self-tolerance, and has been shown to be associated with

diabetes after transplant in a retrospective study on 264 renal transplant recipients (38).

Other immunosuppressive medications such as azathioprine and MMF have not been

associated with diabetes.

b) Obesity: Obesity is an independent risk factor for post-transplant diabetes in adult

recipients, similar to the general population. The risk of diabetes post-transplant was

! 12!

significantly higher in patients with a body mass index (BMI) ≥30 kg/m2 than in patients with a

BMI <30 kg/m2 (relative risk 1.73, 95% CI: 1.57, 1.90) (17). Based on NAPRTCS data, the

prevalence of obesity has increased in children presenting for transplantation, thereby, putting

them at an increased risk of diabetes after transplant. In fact, the prevalence of obesity increased

to 12.4% in children after 1995 as compared to those before 1995 (8%) (39). Obesity causes

diabetes by increasing the peripheral insulin resistance (40).

d) Hypomagnesemia and hypokalemia: In the general population, hypomagnesemia

and hypokalemia have been associated with type 2 diabetes. Low magnesium is considered as a

novel risk factor for diabetes in adult kidney and liver transplant recipients but the results of

interventional studies of supplementation have not demonstrated a significant benefit (41). A

recent meta-analysis showed a pooled relative risk of diabetes of 1.22 (95% CI: 1.09, 1.38) with

post-transplantation hypomagnesemia (42). Hypomagnesemia is linked to defective tyrosine

kinase activation thereby causing impaired insulin-insulin receptor interactions (43, 44) and

reduced insulin secretion (45). Similarly, hypokalemia leads to glucose intolerance by causing

diminished pancreatic beta cell response to glucose. Electrolyte disorders are quite common in

the immediate post-transplant period due to calcineurin inhibitors (CNI), diuretics, and possible

renal tubular dysfunction. However, the literature on the effect of low cations and the risk of

diabetes in pediatric solid organ transplantation is scarce. In a retrospective cohort study

conducted on 451 pediatric solid organ transplant recipients at the Hospital for Sick Children, we

did not find an association of hypomagnesemia and hypokalemia with post-transplant

hyperglycemia and diabetes in the 3 years of follow-up (unpublished data).

e) Growth hormone: The effect of growth hormone on the risk of diabetes has been

controversial. A study by Filler et al. showed that the use of growth hormone pre- and post-

! 13!

transplantation increases the risk of diabetes (46). However, a more recent study did not show

any significant risk of diabetes with the use of growth hormone in children with chronic kidney

disease (47).

2.6 Clinical diagnosis of diabetes mellitus

The criteria for diabetes are defined by the American Diabetes Association guidelines and

are based on three different tests (Table 1) (48). A person has diabetes if a fasting plasma

glucose level is >7.0 mmol/L or a plasma glucose level of >11.1 mmol/L 2 hours after a 75-g

oral glucose tolerance test. Either result must be confirmed by repeat testing on a different day.

The decision of which test is used to diagnose diabetes is left to clinical judgement. Fasting

glucose values between 6.1 to 6.9 mmol/L are defined as impaired fasting glucose (IFG), and 2-

hour plasma glucose values between 7.8 to 11.1 mmol/L are defined as impaired glucose

tolerance (IGT). Both IFG and IGT are important predictors of progression to overt diabetes and

are well-established risk factors for cardiovascular complications (49).

The precise incidence of new onset diabetes after transplantation has been difficult to

determine due to the lack of a standard definition for the condition. Historically, post-transplant

diabetes has been variably defined as having random glucose levels greater than 11.1 mmol/L or

fasting glucose levels greater than 7.8 mmol/L, or the need for insulin or oral hypoglycemic

agents in the post-transplant period. In 2013, an international consensus meeting on post-

transplant diabetes (6) recommended that the definition and diagnosis of diabetes in transplant

recipients should be based on the definition of diabetes mellitus and impaired glucose tolerance

(IGT) described by the American Diabetes Association (Table 1).

! 14!

Table 1 American Diabetes Association criteria for the diagnosis of impaired glucose tolerance and diabetes

Glucose test Normal Impaired Diabetes Fasting plasma glucose

<110 mg/dl or <6.1 mmol/l

100-125 mg/dl or 6.1-6.9 mmol/l (IFG)

>126 mg/dl or >7 mmol/l

2- hr OGTT <140 mg/dl or <7.8 mmol/l

140-199 mg/dl or 7.8-11.1 mmol/l (IGT)

>200 mg/dl or >11.1 mmol/l

Random glucose >200 mg/dl or >11.1 mmol/l + symptoms1

OGTT: oral glucose tolerance test; IFG: impaired fasting glucose; IGT: impaired glucose tolerance; 1symptoms: include polyuria, polydipsia and weight loss 2.7 Use of health administrative data to define diabetes Using administrative data without laboratory information prevent using the ADA

thresholds to define diabetes. However, algorithms have been developed to define diabetes using

administrative data. In Ontario, the health administrative data are used for administering and

monitoring health care delivery and are housed at the Institute of Clinical Evaluative Sciences

(ICES). Multiple algorithms have been developed to accurately identify diabetes using health

administrative data (Table 2). In Ontario, the algorithm uses physician outpatient billing

information through the Ontario Health Insurance Plan (OHIP), as well as data collected from

hospital inpatient charts through the Canadian Institute for Health Information Discharge

Abstract Database (CIHI-DAD). Hux et al. (50) developed a retrospective cohort in adults using

administrative data from hospital discharges and physician’s claims with a diagnosis of diabetes

(ICD-9 250.X). For validation, diagnostic data abstracted from primary care charts (n=3,317) of

57 randomly selected physicians were linked to the administrative data cohort. One hospital

discharge and two physicians claims for diabetes were found to have 86% sensitivity, 97%

specificity, and 80% positive predictive value for defining diabetes.

! 15!

Guttmann et al. (51), described a similar algorithm for identifying diabetes in children. A

total of 700 children were screened from 1994 to 2003 with a prior history of at least one

outpatient or discharge record for diabetes and 300 randomly selected children with no diabetes

records. Both ICD-9 (250.X) and ICD 10 (E10-14) were used. Multiple algorithms were tested,

and it was concluded that four physician claims and no hospital records over a 2-yr period

yielded the most precise definition (83% sensitivity and 99% specificity) of diabetes in children.

However, an important limitation of the algorithm is the inability to differentiate between type 1

and 2 diabetes. There are a number of limitations in using administrative data as it could

underestimate transient diabetes occurring post transplant, gestational diabetes, etc.

Table 2: Studies on the use of health administrative data for identifying diabetes in

children

1ATC: American Therapeutic Classification

2.8 Sequelae after developing diabetes in solid organ transplant recipients

New-onset diabetes mellitus after kidney transplantation is associated with both short-

and long-term complications. In the kidney transplant recipients, new-onset diabetes has been

Author, year Country Data source Code Algorithm

Guttmann, 2010 (51)

Canada Institute of clinical evaluative sciences

ICD 9 and 10 4 physician claim and no hospital record over 2 years

Dabelea, 2009 (52)

USA Indian Health Service Facility

ICD 9 1+ outpatient visit or hospitalization over 3 years

Hsia, 2009 (53) UK Prescription records ATC1 code 1+ prescription claim Cox, 2008 (54) USA Prescription records Antidiabetic

drug 1+ prescription claim

Blanchard, 1997 (55)

Canada Outpatient visits ICD 9 5 outpatient visits or 3–4 if <2 years’ coverage

! 16!

associated with a 2- to 3-fold increased risk of fatal and non-fatal cardiovascular events

compared to non-diabetic recipients (56).

Diabetes post-transplant is also linked to poor patient and graft survival (14). Data from

the United Renal Data System consisting >11,000 kidney transplants between 1996 and 2000

demonstrated that diabetes was associated with a 63% increased risk of graft failure (P <

0.0001), a 46% increased risk of death-censored graft failure (P < 0.0001), and an 87% increased

risk of mortality (P <0.0001) compared to those without diabetes (17).

In pediatric kidney transplant recipients, post-transplant diabetes mellitus has been shown

to be a risk factor for graft failure (RR 1.63, 95% CI: 1.46, 1.84), and increased mortality (RR

1.87, 95% CI: 1.60, 2.18) (57). Diabetes also increases the risk of infections after transplant,

especially sepsis and sepsis-related mortality (58). There is also some evidence that the risk of

CMV infection is also increased in those with diabetes (59).

! 17!

3.! RATIONALE

New-onset diabetes mellitus is one of the main metabolic complications after solid organ

transplantation and has been associated with significant cardiovascular morbidity. It is also a risk

factor for graft failure and increased mortality among transplant recipients (57). It is imperative

to determine the precise incidence of diabetes and timing of its development after transplantation

so that appropriate counselling can be performed and necessary steps taken to reduce or modify

the risk of diabetes and its long-term complications.

Data on the incidence of diabetes mellitus are scarce in pediatric solid organ transplant

recipients and have been previously reported from small centers with poorly defined outcomes.

Moreover, there are few population-based studies with longitudinal follow-up of children post-

transplant once they age into adulthood. It is known that the first few years of transplant confer

the greatest risk of diabetes in recipients; however, the magnitude and direction of the long-term

risk (i.e., up to 10 years) of diabetes in transplanted vs. non-transplanted individuals is unclear.

Also, it is not known whether the risk of diabetes as children transition to adults is similar to the

general population or the risk remains elevated. The impact of recurrent transplants on the risk of

diabetes also has not been explored before.

Our goal is to address this gap in knowledge and evaluate the risk of diabetes mellitus to

better understand and improve long-term outcomes in pediatric solid organ transplants. The

results will yield more precise estimates of the incidence of diabetes thus providing information

to patients and families on long-term risk, the rationale for clinical practice guidelines, and

support data to plan future interventional trials to mitigate risk.

! 18!

4.! OBJECTIVES

Objective 1: To determine the absolute risk, reported as incidence rate per 1000 person-years,

of new-onset diabetes mellitus among children with solid organ transplants between 1991 and

2014 through longitudinal follow-up using validated outcomes from Ontario administrative

healthcare data.

Objective 2: To determine the relative risk of diabetes mellitus among children with solid

organ transplant compared to non-transplanted Ontarian children

Hypothesis: We hypothesize that children with solid organ transplantation are at a higher risk

of diabetes as compared to non-transplanted children

! 19!

5.! METHODS

5.1 Setting and Participants

This is a retrospective cohort study of children who received their first solid organ

transplant at the Hospital for Sick Children, Toronto compared to children who did not get a

transplant and followed longitudinally to determine rates of diabetes. Data from this pediatric

(exposed) cohort and the unexposed cohort are linked to several administrative healthcare

databases available at the Institute for Clinical Evaluative Services (ICES) to determine their

outcomes over long-term follow-up and compared them to healthy children over similar follow-

up (Figure 5). The reporting of the results is in compliance with the RECORD (Reporting of

studies Conducted Using Observational Routinely-collected Data) reporting guidelines (60)

(Appendix 1)

5.2 Study population: Inclusion criteria

1. Exposed cohort: All children (0 to 18 years) who received their first solid organ

transplant (kidney, liver, lung, heart, multiple organs) from 1st January 1991 to 31st December

2014 at the Hospital for Sick Children (n=1020 children). The rationale to include children from

1991 onward is because some of the administrative healthcare databases relevant to this analysis

were unavailable prior to that year.

2. Unexposed cohort: Non-transplanted children chosen from the general population

born in Ontario (estimated ~7 million). The rationale to include all children is that the incidence

of diabetes mellitus in the general population is quite low, but the inclusion of only “healthy”

children may further inflate the relative risk of new-onset diabetes mellitus in transplant

recipients.

! 20!

5.3 Exclusion criteria

1.! Children who are not residents of Ontario at the time of study entry

2.! Individuals aged more than 18 years at the time of study entry

3.! Children with an invalid Ontario health card number (i.e., unable to conduct linkage)

4.! Gap in OHIP eligibility of >1 year

5.4!Data sources and linkage

Transplant database: The transplant database at the Hospital for Sick Children was used to

identify pediatric solid organ transplant recipients between 1991 and 2014. Using names, OHIP

numbers, date of birth, and postal codes, the data from this cohort were linked to health

administrative databases at ICES to determine the study outcomes.

Health administrative databases: ICES databases relevant to this project include the

Ontario Registered Persons Database (RPDB), Ontario Health Insurance Plan (OHIP) Database,

Canadian Institute for Health Information Discharge Abstract Database (CIHI-DAD), and the

Canadian Organ Replacement Register (CORR). These databases were linked using each Ontario

health card number. Every individual was assigned a de-identified ICES key number (IKN), to

maintain patient confidentiality and privacy. The IKNs were used as the patient identifier for all

analysis and linkage. Details of each dataset are provided below:

a) Ontario Health Insurance Plan (OHIP): OHIP physician billing claims contain all

physician billings claimed by physicians, the relevant dates and diagnosis codes. It was used to

identify all outpatient visits for diabetes based on the diagnostic code associated with each visit.

b) The Ontario Registered People Database (RPDB): It contains information such as date of

! 21!

birth, gender, address and date of death. It was used to create the unexposed cohort and to

identify any individuals who died during the study period.

c) The Canadian Institute for Health Information-Discharge Abstract Database (CIHI-

DAD): The CIHI-DAD is a national database that includes information on inpatient

hospitalizations. It was used to identify all children hospitalized for diabetes. Discharge

diagnoses in CIHI-DAD are coded using International Classification of Diseases, Ninth Revision

(ICD-9) until 2002 and the Tenth Revision (ICD-10) after 2002.

d) Canadian Organ Replacement Registry (CORR): CORR is the Canadian national

population-based organ failure registry, and it was used to look back for a history of

transplantation in the unexposed cohort. CORR was not used for creating transplant cohort as it

is known to have missing information on pediatric transplant recipients. On attempting the

linkage of 1106 pediatric transplant cohort with CORR, only 557 children were linked between

1991 and 2014. For this reason, CORR was not used; instead, study cohort was created at

SickKids and later linked to ICES for determining outcomes.

5.5 Outcome assessment and classification: Individuals with diabetes mellitus were determined

over long-term follow-up using previously validated codes and algorithms available in the

Ontario databases held at ICES. Children were followed from the index date until the

development of diabetes mellitus. They were censored at death, transfer out of province, loss of

OHIP coverage, or on 31st March 2015 (Figure 5).

! 22!

Diabetes mellitus: We used a validated definition of diabetes based on ICD (International

Classification of Disease) and OHIP codes determined by Hux et al. on 3317 adults and Guttman

et al. on 700 children as follows.

"! Children (< 18 years): 4 OHIP claims and no hospital records within 2 years were used to

identify children with diabetes mellitus (83% sensitivity, 99% specificity)

"! Adults: 1 hospital or 2 OHIP claims within a 2-year period were used to identify adult-onset

diabetes mellitus among adults (86% sensitivity, 97% specificity; PPV 80%) (50, 51).

The relevant codes to identify patients with diabetes mellitus are as follows:

"! ICD 9: 250.xx "! ICD 10: E10, E11, E 13, E14 "! OHIP Dx: 250.xx "! OHIP fee: Q040, K029, K030, K045, K046

Figure 5: Linkage of pediatric transplant cohort with ICES datasets

Exposed(cohort(n=$1020)

Unexposed(cohort(((((((((((({n=~$7$million}

1. Development$of$diabetes

Reasons$for$censoring1. Transferred$out$of$province2. Death3. Administrative$censoring$31st March$2015

Follow$up$(data$linked$to$ICES)$

Data$source:$RPDBData$source:$CORR,$OHIP,$CIHI$DAD$and$SDS

Data$source:$SickKids$transplant$database

! 23!

5.6 Study timeline

Figure 6 summarizes the study timeline for index date, outcomes and look back window

prior to index date. We used a look back window of 3 years from the index date to exclude any

history of solid organ transplantation in the unexposed group. We didn’t exclude, however,

history of diabetes in the unexposed group prior to inception to the study as the incidence of

diabetes in healthy children is extremely low and may lead to larger reported relative risk if we

excluded those with baseline risk of diabetes.

5.7 Study cohort assembly

Between 1991 and 2014, 1020 children received a solid organ transplant at the Hospital

for Sick Children, and 7,134,067 were included in the unexposed cohort. After applying

exclusion criteria as shown in Table 3, 988 children were retained in transplant cohort and

Look$back(Window

Index&Event&Date

Max(Follow$up(Date

Exposed cohort Unexposed&cohort

Index&event&date Date(of(transplant( Randomly(generated(index(date(

Look&back 3&years:&transplant NA yes

Age&18&

OUTCOMES

1st Jan&1991 31st Dec&2014

Accrual&window

31st Mar&2015

! 24!

5,281,978 in the unexposed cohort. Specifically, around 1.5 million (21%) individuals were

excluded from the unexposed cohort as they had wide gaps in the OHIP eligibility since they

were presumably healthy and rarely visited a health care professional. Hence, individuals with

gaps of more than 1 year were excluded from this study.

Table 3: Exclusion of individuals in exposed and unexposed cohort

IKN: ICES key number

5.8!Covariates: Additional covariates used in the study are defined as follows:

a)! Age at index date: Index date or the date of study entry was designated as the date of

transplant for the exposed cohort, based on the date of transplant at SickKids. The individuals

in the unexposed cohort were those who were born in the same birth year as transplant cohort

and those who survived until inception to the study. An index date was randomly assigned to

them based on the distribution of transplant dates in the exposed group for each birth year

Unexposed Exposed

Criteria Excluded Remaining Excluded Remaining

All6Ontario6residents6(children)6 7134067 1020

Linkage6of exposed6cohort6with CORR ~)500

Perform6the6following6exclusions:1.6Exclude6if6missing6or6invalid6IKN,6DOB6or6sex 0 7134067 0 10202.6Exclude6if6nonKOntario6residents6or6missing6Ontario6status 315037)(4.4%) 6819030 9)(0.9%) 1011

3.6Exclude6if6died6on6or6before6index6date6 3393)(0.05%) 6815637 0 1011

4.6Exclude6if6age6>186on6index6date 25666)(0.3%) 6789971 0 10115.6Exclude6if6history6of6solid6organ6transplant6in6Ontario66outside6SickKids6in6unexposed6only6prior6to61991 215)(0.003%) 6789756 0 1011

8.6Remove6IKNs6not6eligible6at6index 1507778)(21.1%) 5281978 23)(2.2%) 988

FINAL6COHORT 5281978 988

! 25!

and sex. Age at index date was calculated by subtracting index date from the birth date. Age

was treated as a continuous variable.

b)! Sex: Sex was reported from SickKids or ICES as a binary variable.

c)! Neighborhood income quintile: Socioeconomic status (61) was determined using average

neighborhood household income per person, as identified by the subject’s postal code. The

calculated average neighborhood income for each child was categorized into quintiles based

on the average neighborhood income for all persons in Ontario, and was used as a categorical

variable, quintile 1 being the lowest income category ($<40,000) and quintile 5 the highest

($>125,000).

d)! Rural status: The geographic living environment of each individual represented by their

postal code was dichotomized based on the following Statistics Canada definitions: Rural:

areas with a community size of < 10,000 persons, and Urban: all areas outside rural areas

e)! Donor status: Transplant database at SickKids was used to determine information on donor

type, living or deceased for the transplant cohort.

f)! Era: The entire study duration was divided into 2 periods, 1991-2002 and 2003-2014, to

determine whether the risk of diabetes differed in these 2 periods.

5.9 Statistical analysis

All analysis was performed at the ICES Central located at the Sunnybrook Hospital,

Toronto. Continuous variables were reported as mean ± SD or median (IQR) depending on the

distribution of the data. Categorical variables were reported as numbers and percentages.

5.9.1 Analysis for objective 1:

Incidence rates (per 1000 person-years) were calculated overall and at specific time

intervals after study entry (0-1, 1-5, 5-10 and >10 years). The rates were compared and presented

! 26!

as incidence rate ratio (95% CI). Stratified analyses were performed to determine the incidence

rates of diabetes across eras and various organ-specific groups.

5.9.2 Analysis for objective 2:

The Kaplan-Meier product-limit method was used to calculate the cumulative incidence

of diabetes mellitus overall and at 1, 5, 10, and >10 years after study entry. Differences across

survival curves were evaluated using the log-rank test. The proportionality assumption was

graphically examined using log-log plots and scaled Schöenfeld residuals. Due to violations of

the proportional hazards assumption, a time-dependent Cox proportional hazards model was

fitted by dividing follow-up time into periods when the assumption was not violated (i.e., 0-1, 1-

5, 5-10 and >10 years).

Three sequentially nested Cox proportional hazards models were fitted to estimate the

relative hazard of diabetes mellitus in the exposed and unexposed groups after accounting for

potential confounders. Model 1 was the univariate association. Model 2 adjusted for the

demographic variables such as age at index date and sex. Model 3 adjusted for variables in

Model 2 and rural status, income quintile, era and donor status. Stratified analysis was performed

to determine if the risk of diabetes differ by era and across various organ-specific groups.

Any analyses yielding a cell count of 5 or fewer study participants were reported as “< 6”

in accordance with ICES data privacy policies. Analyses were performed using Stata/MP,

version 13 (StataCorp, College Station, TX). A two-sided p-value < 0.05 was considered

statistically significant.

! 27!

5.9.3 Sensitivity analyses

We planned the following sensitivity analyses to check the robustness of the main results.

a)! As lung and multiple organ transplant recipients are at a higher risk of diabetes due to cystic

fibrosis and higher immunosuppressive dosing, we determined the risk of diabetes in exposed

and unexposed groups after excluding lung and multiple organ transplant recipients.

b)! To determine the effect of the first transplant on the risk of diabetes, we censored those

individuals who received an organ transplant after the index date, i.e. individuals with first

organ transplant in the unexposed cohort and those with the second transplant in the exposed

cohort after entering into the study were censored.

5.10 Power calculation

Using the Power Analysis Software (PASS-12), a two-sided log-rank test with a sample

size of 1000 children in the exposed cohort and approximately 5 million in the unexposed cohort,

achieved 90.0% power at a 0.05 significance level to detect a hazard ratio as low as 1.99 when

the proportion of diabetes in the unexposed group was 4%. We assumed a drop off rate of 1% in

unexposed group and 5% in the exposed group (Figure 7).

! 28!

Figure 7: Log-rank test analysis showing the detectable hazard ratios for diabetes in the

study cohort of transplanted and non-transplanted children

5.11 Research Ethics Approval

Institutional ethics approval was obtained from the SickKids, and additional administrative

ethics approval was obtained from the University of Toronto. A data sharing agreement was

created between SickKids and ICES central to allow data transfer to ICES. The datasets were

linked using unique encoded identifiers and analyzed at the Institute for Clinical Evaluative

Sciences (ICES). This study was approved by the institutional review board at Sunnybrook

Health Science Centre, Toronto, Canada.

! 29!

6! RESULTS

6.1 Study cohort

Based on the inclusion and exclusion criteria discussed above, 988 individuals were

retained in transplant cohort and 7,134,067 in the unexposed cohort.

Figure 8:!Study cohort after meeting the inclusion and exclusion criteria!

Exposed(cohort((Derived(from(SickKids)

N=1020 N=7,134,067

Exposed(cohort:(N=988Kidney:(N=406Heart:(N=221Liver:(N=310Lung:(N=37

Multiple and(small(bowel: N=14

Unexposed(cohort(N=5,281,978

Excluded(N=1,852,089

Excluded(N=32

Unexposed(cohort((Derived(from(ICES)

Linked(to(ICES

! 30!

6.2 Baseline characteristics of the exposed and unexposed cohort

Baseline features, such as age, sex, and socio-economic status, were comparably

distributed across both cohorts. The mean age at index date was approximately 8 years in both

groups (Table 4). The median follow-up duration was 9.2 (IQR: 3.9, 15.9) years for the entire

study cohort.

Table 4: Baseline characteristics of the exposed and unexposed cohort!

Demographics n Exposed1cohort n Unexposed1cohort

Mean1age1(SD),1years 988 7.8± 6.2 5,281,978 8.2± 6.4Males,1n1(%) 988 536-(54.2) 5,281,978 2,632,984-(49.8)Era,1n1(%)1991>2002 988 398-(40.3) 2,224,510-(42.1)2003>2014 590-(59.7) 3,057,468-(57.9)

Income1quintile,1n1(%) 986 5,245,1011 216-(21.9) 1,077,103-(20.4)2 195-(19.7) 1,012,407-(19.2)3 199-(20.1) 1,047,173-(19.8)4 207-(20.9) 1,091,174-(20.7)5 171-(17.3) 1,016,244-(19.2)

Rural1status, n1(%) 137-(13.9) 688,941-(13.1)

! 31!

6.3 Baseline characteristics of organ-specific pediatric transplant recipients

Baseline features in the individual transplant groups are highlighted in Table 5. Of 988

transplants (54.2% males), kidney and liver organ groups comprised 70% of all transplants

followed by heart, lung and multiple organ groups. Due to small number of patients in each

group, the lung and multiple organ recipients are presented as one group. Kidney, lung and

multiple organ recipients were older compared to liver and heart recipients at the time of

transplant. There were no significant differences among various organ groups except the number

of individuals in both eras and living donation rates across various organ groups.

Table 5: Baseline characteristics of organ-specific pediatric transplant recipients

Kidney Heart LiverLung/and/Multiple Overall

(n=406) (n=(221) (n=310) (n=51) (n=(988)Mean/age/at/transplant/(years) 10.7± 5.0 5.7± 6.3 4.9± 5.7 10.4± 5.7 7.8± 6.2

Males,/n/(%) 239((58.9) 121((54.7) 155((50.0) 21((41.2) 536((54.2)

Era/of/transplant,/n/(%)1991>2002 182((44.8) 89((40.3) 115((37.1) 12((23.5) 398((40.3)

2003>2014 224((55.2) 132((59.7) 195((62.9) 39((76.5) 590((59.7)

Living/donation,/n/(%) 178((43.8) 0 97((31.4) 0 275((27.9)

Income/quintile,/n/(%)1 78((19.2) 49((22.2) 81((26.1) 8((15.7) 216((21.9)

2 90((22.2) 40((18.1) 57((18.4) 8((15.7) 195((19.7)

3 82((20.2) 51((23.1) 52((16.8) 14((27.4) 199((20.1)

4 84((20.7) 40((18.1) 72((23.2) 11((21.6) 207((20.9)

5 72((17.7) 41((18.5) 48((15.5) 10((19.6) 171((17.3)

! 32!

6.4 Incidence rate of diabetes in the exposed and unexposed cohort

During 5,601,9824 person-years of follow-up, the incidence rate of diabetes in the

exposed cohort (IR 17.8 per 1000 person-years, 95% CI: 15, 21) was significantly higher as

compared to the unexposed cohort (IR 2.5 per 1000 person-years, 95% CI: 2.5, 2.5) (Table 6).

Compared to the unexposed group (141,108, 2.7%), the number of individuals with diabetes was

higher in the exposed group (138, 14%). The median age at onset of diabetes was also

significantly lower in the exposed group (15.7 years, IQR: 13.3, 19.9) as compared to unexposed

group (21.4 years, IQR: 15.6, 27.8). Moreover, the transplanted children developed diabetes

much earlier (1.9 years, IQR: 0.3, 8.6) than the unexposed cohort (8.7 years, IQR: 3.6, 14).

Ninety-eight (71%) individuals in the exposed cohort developed diabetes at < 18 years of age

compared to 49,635 (35%) in unexposed cohort.

The incidence rate of diabetes (61.6 per 1000 person-years, 95% CI: 47.2, 80.4) of

diabetes was the highest within the first year after transplantation. After the first year, the

incidence rate declined to one-fifth but remained consistently 5-6 times greater than those in the

unexposed cohort even after a decade of follow-up.

! 33!

Table 6: Incidence rate for diabetes in the exposed and unexposed cohort overall and at

specific time intervals

Overall Exposed cohort Unexposed cohortNumber'with'diabetes,'n'(%) 138'(14%) 141108'(2.7%)IR (per'1000'person@years) 17.8'(15.0,'21.0) 2.5'(2.5,'2.5)Incidence rate'ratio'(95%CI)* 7.0'(5.9,'8.3) Ref.'0315yearsNumber'with'diabetes 54 15844IR'(per'1000'person@years) 61.6'(47.2,'80.4) 3.1'(3.0,'3.1)Incidence rate'ratio'(95%CI)* 11.9'(9.0,'15.6) Ref.'1355yearsNumber'with'diabetes 35 29069IR'(per'1000'person@years) 12.8'(9.2,'17.8) 1.7'(1.7,'1.7)Incidence rate'ratio'(95%CI)* 5.8'(4.1,'8.1) Ref.'53105yearsNumber'with'diabetes 22 34538IR'(per'1000'person@years) 9.7'(6.4,'14.8) 2.1'(2.1,'2.2)Incidence rate'ratio'(95%CI)* 3.2'(2.0,'4.8) Ref.'>105yearsNumber'with'diabetes 27 61657IR'(per'1000'person@years) 14.3'(9.8,'20.8) 3.5'(3.5,'3.5)Incidence rate'ratio'(95%CI)* 3.6'(2.4,'5.3) Ref.'*'p'value<0.0001

! 34!

6.5 Stratified analysis of incidence rate of diabetes in the exposed and unexposed cohort, by

era

In both eras, the risk of diabetes in the exposed cohort was significantly higher than the

unexposed cohort. In the first year, the incidence rate (69.1 per 1000 person-years, 95% CI: 49.9,

95.8) of diabetes was higher in the era 2 (2003- 2014) for the exposed group as compared to era

1 (1991- 2002) (50.5 per 1000 person-years, 95% CI: 31.8, 80.2). This trend continued until 5

years after the index date (Table 7).

Table 7: Incidence rate for diabetes in exposed and unexposed cohort at specific time

intervals, stratified by era

1991#2002 2003#2014Exposed Unexposed Exposed Unexposed

OverallNumber'with'diabetes 78 112197 60 28911IR'(per'1000'person<years) 15.7'(12.6,'19.6) 2.9'(2.8,'2.9) 21.4'(16.6,'27.5) 1.7'(1.7,'1.7)Incidence rate'ratio'(95%'CI)* 5.5'(6.8,'14.2) Ref.' 12.5'(16.1,'4.8) Ref.'0#16yearsNumber'with'diabetes 18 7403 36 8441IR'(per'1000'person<years) 50.5'(31.8,'80.2) 3.3'(3.3,'3.4) 69.1'(49.9,'95.8) 2.9'(2.8,'2.9)1#56yearsNumber'with'diabetes 16 17252 19 11817IR'(per'1000'person<years) 12.3'(7.5,'20.0) 1.9'(1.9,'2.0) 13.2'(8.4,'20.7) 1.4'(1.4,'1.4)5#106yearsIR'(per'1000'person<years) 12.2'(7.7,'19.3) 2.4'(2.4,'2.5) 5.1'(1.9,'13.7) 1.5'(1.5,'1.6)>106yearsIR'(per'1000'person<years) 14.2'(9.7,'20.9) 3.5'(3.5,'3.6) 15.9'(2.2,'112.9) 1.8'(1.7,'2.0)*p value<0.0001

! 35!

6.6 Stratified analysis of the incidence rate of diabetes, by individual organ groups

Among the individual organ groups, the incidence rate of diabetes (per 1000 person-

years) was highest in the lung and multiple organ transplant recipients (IR 111.3, 95% CI: 70.7,

178.2). The risk was highest within the first year of transplant and was 5 to 20 times higher

compared to other organ recipients. Risk declined after the first year but remained 2 to 4 times

higher compared to other organ recipients until a decade after transplant (Table 8).

Table 8: Incidence rate of diabetes, overall and at specific time intervals, stratified by

organ groups

Kidney Heart Liver Lung/&/multipleOverallNumber'with'diabetes 70'(17.2) 26'(11.8) 24'(7.7) 18'(35.3)Person'years'of'observation 3718.8 1503.3 2386.9 160.3

IR (per'1000'personEyears) 18.8'(14.9,'23.8) 17.3'(11.8,'25.4) 10.0'(6.7,'15.0) 111.3'(70.7,'178.2)061/yearsNumber'with'diabetes 28 6 6 14Person'years'of'observation 376.3 187.5 279.1 33.9IR'(per'1000'personEyears) 74.4'(51.4,'107.8) 32.0'(14.4,'71.2) 21.5'(9.6,'47.8) 412.2'(244.1,'696.1)

165/years

IR'(per'1000'personEyears) 12.3'(7.4,'20.3) 14.9'(7.8,'28.7) 9.4'(4.7,'18.7) 48.0'(15.5,'148.8)

5610/years

IR'(per'1000'personEyears) 11.9'(6.9,'20.6) 6.6'(2.1,'20.5) 7.4'(3.1,'17.8) 23.3'(3.3,'165.6)

>10/years

IR'(per'1000'personEyears) 13.6'(8.0,'22.9) 30.8'(15.4,'61.6) 8.6'(3.6,'20.7) EE

! 36!

6.7 Relative risk of diabetes in exposed and unexposed cohort

The cumulative incidence of diabetes at 1, 5, 10, and >10 years in the exposed cohort was

5.8%, 10.3%, 14.8% and 36% compared to 0.3%, 0.9%, 2.0% and 6.7% in the unexposed cohort.

Overall, the exposed cohort had 7.5 times (HR 7.4, 95% CI: 6.2, 8.7) higher relative hazard of

diabetes as compared to the unexposed cohort. After adjusting for potential confounders, there

was no significant change in the relative hazard (HR 8.9, 95% CI: 7.5, 10.5). The adjusted

relative hazard of diabetes was highest in the first year (HR 20.7, 95% CI: 15.9, 27.1) after study

entry. Later, the relative hazard of diabetes reduced but remained approximately 5 to 8 times

higher in the exposed cohort as compared to the unexposed cohort (Figure 9 and Table 9).

Figure 9: Kaplan-Meier curve for cumulative risk of diabetes in exposed and unexposed cohort

! 37!

Table 9: Hazard ratio for diabetes in exposed and unexposed cohort overall and at specific

time intervals

1Model 1: unadjusted model; 2Model 2: adjusted for age at study entry and sex; 3Model 3: adjusted for age, sex, income quintile, rural status, era of transplant

Overall Exposed-cohort Unexposed-cohortModel&1 (HR,&95%&CI) 7.4&(6.2,&8.7) Ref.Model&2 (HR,&95%&CI) 8.9&(7.5,&10.5) Ref.Model&3 (HR,&95%&CI) 8.9&(7.5,&10.5) Ref.041-yearsModel&1 19.7&(15.1,&25.7) Ref.Model&2 20.9&(15.9,&27.2) Ref.Model&3 20.7&(15.9,&27.1) Ref.145-yearsModel&1 7.6&(5.4,&10.6) Ref.Model&2 8.3&(6.0,&11.6) Ref.Model&3 8.4&(6.0,&11.7) Ref.5410-yearsModel&1 4.5&(2.9,&6.9) Ref.Model&2 5.5&(3.6,&8.4) Ref.Model&3 5.6&(3.7,&8.4) Ref.>10-yearsModel&1 4.2&(2.9,&6.1) Ref.Model&2 5.6&(3.8,&8.2) Ref.Model&3 5.6&(3.8,&8.2) Ref.

! 38!

6.8 Stratified analysis of the risk of diabetes in exposed and unexposed groups by era

The relative hazard of diabetes, after adjusting for potential confounders, was higher in

the exposed cohort during years 2003 to 2014 (HR 12.6, 95% CI: 9.8, 16.3) compared to the

hazard during years 1991 to 2002 (HR 7.3, 95% CI: 5.8, 9.1) (Table 10 and Figure 10a and b).

! 39!

Figure 10: Cumulative risk of diabetes in exposed and unexposed cohort in era 1(1991-

2002, a) and era 2, (1993-2014, b)

a)!

!!!

b)

! 40!

Table 10: Hazard ratio for diabetes in exposed and unexposed cohort at specific time

intervals, stratified by era

1Model 1: unadjusted model; 2Model 2: adjusted for age at study entry and sex; 3Model 3: adjusted for age, sex, income quintile, rural status

1991#2002 2003#2014Exposed Unexposed Exposed Unexposed

OverallModel&1 (HR,&95%&CI) 5.7&(4.6,&7.1) Ref. 12.3&(9.6,&15.9) Ref.Model&2&(HR,&95%&CI) 7.3&(5.8,&9.1) Ref. 12.6&(9.8,&16.3) Ref.Model&3 (HR,&95%&CI) 7.3&(5.8,&9.1) Ref. 12.6&(9.8,&16.3) Ref.0#16yearsModel&1 14.9&(9.4,&23.7) Ref. 23.6&(17.0,&32.8) Ref.Model&2 17.5&(11.0,&27.9) Ref. 22.4&(16.2,&31.2) Ref.Model&3 17.6&(11.0,&27.9) Ref. 22.5&(16.3,&31.3) Ref.1#56yearsModel&1 6.3&(3.8,&10.3) Ref. 9.4&(6.0,&14.8) Ref.Model&2 7.4&(4.5,&12.1) Ref. 9.6&(6.1,&15.0) Ref.Model&3 7.4&(4.6,&&12.2) Ref. 9.6&(6.1,&15.0) Ref.5#106yearsModel&1 5.0&(3.1,&7.9) Ref. 3.3&(1.2,&8.9) Ref.Model&2 6.3&(3.9,&9.9) Ref. 3.6&(1.4,&9.7) Ref.Model&3 6.3&(4.0,&10.0) Ref. 3.7&(1.4,&9.8) Ref.>106yearsModel&1 4.1&(2.8,&6.0) Ref. 8.7&(1.2,&61.6) Ref.Model&2 5.5&(3.8,&8.1) Ref. 9.7&(1.4,&68.8) Ref.Model&3 5.5&(3.8,&8.1) Ref. 10.1&(1.4,&71.8) Ref.

! 41!

6.9 Risk of diabetes in individual pediatric solid organ transplant groups

Among the individual organ groups, lung and multiple organ groups were at the highest

risk of developing diabetes. The relative hazard of diabetes was 5 times higher (HR 5.4, 95%

CI:3.0, 9.8) compared to kidney transplant recipients during the entire study period.

The risk of developing diabetes in lung and multiple organ recipients was highest within

the first year after transplant (HR 5.6, 95% CI: 2.6, 12.3). The risk in other organ groups was not

significantly different from kidney organ recipients in the first year of transplant. After the first

year of transplant, the risk in the lung and multiple organ groups reduced but stayed 2 to 5 times

higher compared to kidney transplant recipients.

After a decade of transplant, the relative hazard of diabetes was maximum in the heart

transplant recipients compared to kidney transplant recipients after adjusting for potential

confounders (HR 3.6, 95% CI: 1.2, 10.6) (Figure 11 and Table 11).

! 42!

Figure 11: Cumulative risk of diabetes in individual pediatric organ transplant recipients

! 43!

Table 11: Hazard ratio for diabetes in individual pediatric solid organ groups overall and

at specific time intervals after index date

1 Model 1: unadjusted model;

2Model 2: adjusted for age at study entry and sex;

3Model 3: adjusted for age, sex,

income quintile, rural status, era of transplant and donor status

Kidney Heart Liver Lung/&/multipleOverallModel&11 (HR,&95%&CI) Ref. 0.9&(0.6,&1.4) 0.5&(0.3,&0.8) 4.4&(2.6,&7.5)Model&22 (HR,&95%&CI) Ref. 1.5&(0.9,&2.4) 0.9&(0.6,&1.5) 5.1&(3.0,&8.7)Model&33 (HR,&95%&CI) Ref. 1.6&(0.9,&2.6) 0.9&(0.6,&1.6) 5.4&(3.0,&9.8)061/yearsModel&1 Ref. 0.4&(0.2,&1.0) 0.3&(0.1,&0.7) 4.9&(2.6,&9.3)Model&2 Ref. 0.8&(0.3,&2.0) 0.6&(0.2,&1.5) 5.1&(2.7,&9.9)Model&3 Ref. 1.0&(0.4,&2.5) 0.7&(0.3,&1.7) 5.6&(2.6,&12.3)165/yearsModel&1 Ref. 1.2&(0.5,&2.8) 0.7&(0.3,&1.8) 3.8&(1.1,&13.0)Model&2 Ref. 2.1&(0.9,&4.8) 1.5&(0.6,&3.7) 5.0&(1.4,&17.5)Model&3 Ref. 2.2&(0.8,&5.8) 1.6&(0.6,&3.9) 5.3&(1.4,&20.2)5610/yearsModel&1 Ref. 0.6&(0.2,&2.0) 0.6&(0.2,&1.8) 2.1&(0.3,&15.8)Model&2 Ref. 0.9&(0.2,&3.5) 1.1&(0.3,&3.2) 2.8&(0.4,&22.0)Model&3 Ref. 0.7&(0.2,&2.9) 0.9&(0.3,&2.7) 2.0&(0.2,&17.3)>10/yearsModel&1 Ref. 2.7&(1.1,&6.6) 0.7&(0.2,&2.0) ;Model&2 Ref. 3.2&(1.2,&8.4) 0.8&(0.3,&2.4) ;Model&3 Ref. 3.6&(1.2,&10.6) 0.7&(0.2,&2.3) ;

! 44!

6.10 Transplant in exposed and unexposed cohort after study index date

Among individuals in the unexposed cohort, 1,139 individuals received an organ

transplant after the index date. Of the 1139 patients, 798 (70%) received a kidney, 46 (4.0%)

heart, 141 (12.4%) liver, 114 (10.0%) lung and 40 (3.5%) multiple organ transplantation.

Moreover, 231 individuals received a second transplant and 37 received more than 2 transplants.

The median duration to first transplant was 10.4 years (IQR: 5.6, 15.1) and the median age at

transplant was 23.4 years (IQR: 19.2, 28.9). Among those who underwent a transplant, 123

(10.8%) developed diabetes after the transplant. Of note, 97 (78.9%) developed diabetes after the

first transplant and the rest after subsequent transplants.

Among the children in the transplant group, 637 (64.5%) children received a subsequent

transplant. Of them, 83 (13.0%) received a third transplant. Kidney (n=292) and liver (n=209)

were the most common second transplants. Of 138 with diabetes, 53 (38.4%) developed it after

1st organ transplant and 85 (61.6%) developed it after the 2nd transplant.

! 45!

6.11Sensitivity analysis

6.11.1 Risk of diabetes in exposed and unexposed cohort after excluding lung and multiple

organ transplant recipients

The risk of diabetes in the exposed cohort was 8 times (HR 7.9, 95% CI: 6.6, 9.5) higher

than the unexposed cohort after excluding lung and multiple organ transplant recipients. Within

the first year, the risk was 16 times higher (HR 16.2, 95% CI: 11.9, 22.1) after adjusting for

potential confounders. The estimates did not differ significantly from the primary analysis

(Figure 12 and Table 12).

Figure 12: Cumulative risk of diabetes in exposed and unexposed cohort after excluding

lung and multiple organ recipients

! 46!

Table 12: Risk of diabetes in exposed and unexposed cohort after excluding lung and

multiple organ transplant groups

Overall Exposed-cohort Unexposed-cohortModel&1 (HR,&95%&CI) 6.6&(5.5,&7.8) Ref.Model&2 (HR,&95%&CI) 7.9&(6.6,&9.4) Ref.Model&3 (HR,&95%&CI) 7.9 (6.6,&9.5) Ref.041-yearsModel&1 15.2&(11.2,&20.8) Ref.Model&2 16.3&(11.9,&22.2) Ref.Model&3 16.2&(11.9,&22.1) Ref.145-yearsModel&1 7.1&(5.0,&10.0) Ref.Model&2 7.8&(5.5,&11.0) Ref.Model&3 7.9&(5.8,&11.1) Ref.5410-yearsModel&1 4.4&(2.9,&6.8) Ref.Model&2 5.3&(3.5,&8.2) Ref.Model&3 5.4&(3.5,&8.3) Ref.>10-yearsModel&1 4.2&(2.9,&6.2) Ref.Model&2 5.7&(3.9,&8.2) Ref.Model&3 5.7&(3.9,&8.3) Ref.

! 47!

6.11.2 Risk of diabetes in exposed and unexposed cohort after censoring individuals at

transplant after the index date

The relative hazard of diabetes was 22 times higher in exposed group compared to

unexposed group in the first year after adjusting for potential confounders. The risk remained

elevated to 4 times even after 5 years after index date. The results were similar to those seen in

the primary analysis above (Table 13).

Table 13: Risk of diabetes in exposed and unexposed cohort after censoring individuals at

transplant after the index date

1 Model 1: unadjusted model; 2Model 2: adjusted for age at study entry and sex; 3Model 3: adjusted for age, sex, income quintile, rural status, and era of transplant

Overall Exposed-cohort Unexposed-cohortModel&11 (HR,&95%&CI) 10.4 (8.0,&13.6) Ref.Model&22 (HR,&95%&CI) 12.5&(9.5,&16.4) Ref.Model&33 (HR,&95%&CI) 12.8&(9.8,&16.8) Ref.041-yearsModel&1 20.1&(14.1,&28.5) Ref.Model&2 22.3&(15.7,&31.8) Ref.Model&3 22.1(15.5,&31.4) Ref.145-yearsModel&1 9.9&(6.0,&16.1) Ref.Model&2 11.1&(6.8,&18.0) Ref.Model&3 11.6 (7.1, 18.9) Ref.>5-yearsModel&1 3.1&(1.4, 7.0) Ref.Model&2 4.1&(1.9,&9.2) Ref.Model&3 4.4&(2.0,&9.9) Ref.

! 48!

6.11.3 Risk of diabetes among organ groups after censoring individuals at transplant after

the index date

The relative hazard of diabetes in lung and multiple organ transplant recipients remained

8 times higher than kidney transplant recipients during the entire study period after censoring

those who received a second organ transplant after the index date, similar to risk in the primary

analysis (Table 14).

Table 14: Risk of diabetes in organ groups after censoring individuals at transplant after

the index date

1 Model 1: unadjusted model; 2Model 2: adjusted for age at study entry and sex; 3Model 3: adjusted for age, sex, income quintile, rural status, era of transplant and donor status !

Kidney Heart Liver Lung/and/MultipleOverallModel&1 (HR,&95%&CI) Ref. 0.7&(0.4,&1.4) 0.2&(0.0,&0.8) 6.9&(3.5,&13.9)Model&2 (HR,&95%&CI) Ref. 1.2&(0.6,&2.3) 0.3&(0.1,&1.4) 6.4&(3.2,&19.6)Model&3 (HR,&95%&CI) Ref. 1.3&(0.5,&3.2) 0.4&(0.1,&1.7) 7.9&(3.2, 19.7)051/yearsModel&1 Ref. 0.6&(0.2,&1.6) 0.2&(0.0,&1.1) 8.2&(3.5,&18.9)Model&2 Ref. 1.1&(0.4,&3.1) 0.5&(0.1,&2.5) 7.5&(3.2,&17.5)Model&3 Ref. 2.0&(0.5,&7.9) 0.8&(0.2,&4.2) 13.5&(3.7,&49.7)155/yearsModel&1 Ref. 0.7&(0.2,&2.0) ; 3.5&(0.7,&17.0)Model&2 Ref. 1.1&(0.4,&3.2) ; 3.5&(0.7,&17.0)Model&3 Ref. 0.7&(0.2,&2.3) ; 2.2&(0.4,&12.5)>/5 yearsModel&1 Ref. 1.2&(0.2,&6.5) ; ;Model&2 Ref. 0.9&(0.1,&6.2) ; ;Model&3 Ref. 1.3&(2.0,&8.6) ; ;

! 49!

7.! DISCUSSION

This is the first study to determine the long-term risk of new-onset diabetes mellitus in

pediatric solid organ transplant recipients compared to the non-transplanted children in Ontario

using validated algorithms through the health administrative data as children age into adulthood.

Our results demonstrate that 70% of the transplant recipients develop diabetes during childhood

(<18 years), and the risk of new-onset diabetes is highest within the first year of transplant.

During the entire study period, the overall risk of diabetes is 9 times higher in solid organ

transplant recipients compared to non-transplanted children. Within the first year, however, the

risk is 20 times higher in those with a solid organ transplant compared to healthy children. After

the first year, the risk of diabetes in solid organ recipient's declines but remains 5 to 8 times

higher even a decade after transplant compared to non-transplanted individuals. Among the

individual organ groups, lung and multiple organ transplant recipients have the highest risk of

having diabetes compared to heart, kidney or liver organ transplant recipients. About two-thirds

of solid organ transplant recipients also receive a second transplant, and the risk of diabetes is

even higher after the second compared to the first transplant. We demonstrate that diabetes is a

considerable burden for children after transplantation and remains an issue as children age into

young adults.

No prior studies have compared the risk of diabetes to a healthy pediatric cohort, nor

followed children after age 18 years to determine the risk as young adults. There are few

important findings in our study that deserve mention. First, a majority of individuals with

transplant develop diabetes during their childhood in their mid-teens when they are in the process

of being transitioned to adult nephrology care. Hence, it is important to screen these children for

diabetes during the peri-transition period so that appropriate management can be provided in

! 50!

time. This approach will not only lead to early detection and treatment but also prevent or delay

the long-term complications associated with diabetes such as cardiovascular disease. Second,

solid organ transplant recipients continue to remain at a 5-times higher risk of diabetes compared

to non-transplanted individuals even after 10 years' post-transplant despite being on low

immunosuppression, thereby, highlighting the need for continuous monitoring of blood glucose

levels. This is in contrast to the current understanding that the risk of diabetes in organ recipients

becomes similar to the general population after a few years of transplant. Third, the risk of

diabetes increases with recurrent transplants. This suggests that close monitoring before and after

subsequent transplantation is warranted to ensure appropriate dosing and choice of various

diabetogenic immunosuppressive medications and early screening for diabetes to minimize

cardiovascular and infectious complications.

Diabetes mellitus is a major public health concern, especially in children and youth as the

prevalence has increased remarkably over the last decade. Similar to the trends in the general

population, the burden of diabetes has escalated among pediatric solid organ recipients. In fact,

in 2001, data from the North American Pediatric Renal Transplant Cooperative Study

(NAPRTCS) on 1365 recipients reported an incidence of diabetes of only 3% (9). Later, studies

reported an incidence of new-onset diabetes as high as 8 to 20% in pediatric kidney transplant

recipients (57). In pediatric liver transplant recipients, the published literature suggests that the

risk of diabetes is around 8 to 14% (62). However, the risk in some disease groups such as cystic

fibrosis and primary sclerosing cholangitis have been reported as high as 20 to 50% (63). Data

on the risk of diabetes in pediatric heart, lung and multiple organ recipients are sparse. A study

on 78 heart and heart/lung transplant recipients published around 2 decades ago showed an

incidence of diabetes as high as 43% (13). Yet, the incidence in heart recipients was only 4%.

! 51!

Our findings are consistent with the published literature as we demonstrate the highest incidence

of diabetes in lung and multiple organ transplant recipients. There are a number of factors that

can explain the significant discrepancy in the burden of diabetes within organ groups. The

definition of diabetes has not been consistent across studies nor the timing of diagnosis of

diabetes after transplant. Moreover, the variable follow-up time after the transplant is also

responsible for the differences in the reported incidence of diabetes among various organ groups.

To address these limitations, we used validated algorithms to define diabetes across a large

cohort of solid organ transplant recipients and demonstrated that the risk of diabetes in maximum

within the first year of transplant and lung and multiple organ transplant recipients are at the

highest risk of diabetes compared to other organ groups.

Health administrative data provide reliable estimates of the disease burden at a population

level and are quite useful to understand the trend in various diseases. A number of studies in

children have used health administrative databases to evaluate diabetes burden by using various

algorithms. Blanchard et al. (55) used administrative data in Manitoba between 1985-1993 to

evaluate type 1 diabetes in children <14 years of age. The case definition used was five or more

physician claims or a minimum of three physician claims if registered with Manitoba Health for

less than 2 years. The overall ascertainment rate of the Manitoba Diabetes Database was 95% for

incident cases and 93% for prevalent cases.!Rhodes et al. (64) used a Boston Massachusetts

database to evaluate ICD-9 codes in children with diabetes. The positive predictive value was

97% for type 1 diabetes using these codes. Studies have shown that use of two or more physician

claims and 1-2 years of data in the algorithm is more accurate than using only 1 outpatient code

to identify individuals with diabetes (65). Very few studies have included drug claims in the

algorithm to identify diabetes. In Ontario, health administrative databases don’t capture

! 52!

information on medications in individuals under 65 years of age. However, Dart et al. (65) used

health administrative data at Manitoba and showed that addition of prescription claims to the

algorithm didn’t change the sensitivity of identifying diabetes significantly. We used separate

validated algorithms for children and adults with >95% specificity to identify diabetes in Ontario

among transplanted and non-transplanted individuals.

We also compared the risk of diabetes in different eras and showed that the risk was

higher in the most recent era (2003- 2014). This finding can be explained by a number of factors,

such as temporal trends in the clinical care. There has been a significant change in the post-

transplant care, lifestyle and dietary habits, and the choice of immunosuppressive medications

across all organ groups. In particular, there has been a surge in the use of tacrolimus over the last

decade in the transplant recipients which is more diabetogenic than cyclosporine. Also, there has

been an increase in the obesity both in the general and the transplant population (39) which may

be responsible for the higher incidence of diabetes recently. Second, there has been an increase

in the burden of CMV infection which is associated with diabetes (14) and may be a potential

reason for the difference in the risk of diabetes in the two eras. Finally, the higher incidence of

diabetes in the more recent era may be due the bias caused by a higher awareness of diabetes

post-transplant by physicians over the last decade leading to more billing claims.

We also adjusted for important confounders such as income quintile and rural status

which did not significantly attenuate the association. Neighborhood income quintile is an indirect

measure of the socioeconomic status that can impact the development of diabetes by influencing

on the health behavior, lifestyle, and dietary habits. An important limitation of health

administrative data is the inability to get individual-level information on socioeconomic status.

! 53!

7.1 Strengths of the study

We studied the long-term risk of developing diabetes in a large pediatric solid transplant

cohort compared with the non-transplanted individuals in Ontario and were able to provide

precise estimates of diabetes across all organ groups through validated algorithms using health

administrative databases. We utilize the strengths of a single large regional referral center and

universal health care system that captures over 95% of transplants in Ontario.!Use of health

administrative data also provides remarkable power and the ability to follow patients over a

prolonged period of time. This study addresses prior concerns of predominantly cross-sectional

studies, small sample sizes or short follow-up. It is novel as there are few studies among liver,

lung and heart recipients with long-term outcomes beyond age 18.

7.2 Limitations

Our study has several limitations that deserve note. First, health administrative data do

not contain information on several clinically relevant risk factors for diabetes, such as

immunosuppressive medications, laboratory data on drug levels, family history of diabetes,

lifestyle and dietary habits, which could potentially contribute to the development of diabetes but

could not be accounted for in the analyses. Second, since health administrative data are being

used to identify individuals with diabetes through validated algorithms, it is possible that some

individuals with the outcome of interest are not captured as the sensitivity of the algorithm to

identify children and adults with diabetes are only 83% and 86%, respectively. However, any

misclassification bias caused by imperfect sensitivity will likely be non-differential with respect

to the exposure status. Another limitation is that some cases of medication induced diabetes or

other forms of transient hyperglycemia such as gestational diabetes may have been inadvertently

! 54!

included in the exposed cohort. Third, a potential source of misclassification bias is the increased

surveillance of the transplant recipients. Children with a solid organ transplant frequently visit

the health care facility and are more likely to be diagnosed with diabetes as compared to healthy

children who don't visit hospitals that often. A potential way to eliminate this bias will be

including a cohort of children, as controls, with a condition such as asthma which may warrant

frequent visits to a physician. However, these children are also on prednisone during their

disease exacerbation, and therefore won’t be appropriate as “healthy” controls. Fourth, we didn't

rule out diabetes in the unexposed cohort prior to the study entry. However, as our data show, the

incidence of diabetes in the general population is extremely low thus, there will very few

individuals with pre-existing diabetes in the unexposed cohort. Fifth, we did not account for the

repeating transplant events and censored at the time of second transplant. Sixth, through health

administrative data we couldn't differentiate between type 1 and type 2 diabetes. However, based

on the published evidence, the pathophysiology of new-onset diabetes is similar to type 2

diabetes. Finally, due to privacy issues, we merged the lung and multiple organ transplant

recipients. These organ groups may have different mechanism and risk of diabetes and

presenting them as a single group might have biased our estimates.

Despite these limitations, we believe that our study provides strong evidence to support

regular screening of diabetes in children with solid organ transplantation.

7.3 Implications of the study

Despite improvements in survival immediately post-transplant, there is a significant

morbidity and mortality among childhood solid organ recipients as they age into adulthood.

New-onset diabetes is an important risk factor for short- and long-term complications in

! 55!

transplant patients. With increasing life expectancy after pediatric transplantation and recurrent

transplants over their lifetime, it is important to study diseases such as diabetes that develop as

children age.

Our results have provided more precise and consistent estimates of the burden of diabetes

in a large cohort of transplant recipients. This information will better inform the parents and

families about the long-term risk of this important post-transplant complication. Currently, the

level of evidence in the existing guidelines on the blood glucose monitoring in the early and late

post-transplant period is quite poor (2D). Kidney Disease Improving Global Outcomes (KDIGO)

recommends blood glucose monitoring every week for the first 4 weeks after transplant, and

every 3 months for the first year, and yearly thereafter in the adult kidney transplant recipients

(6, 66). However, no such guidelines exist for children and young adults with a solid organ

transplant. Our results provide a rationale for clinical practice guidelines on screening in children

for diabetes not only in the early post-transplant period but also few years’ post-transplant.

7.4 Conclusions

Children with solid organ transplantation are a higher risk of diabetes compared to non-

transplanted individuals. The risk is highest in the first year of transplant but remains elevated

even after a decade after transplantation. Specifically, transplanted children are at 20 times

higher risk of developing diabetes in the first year, and at 5 times higher risk thereafter. Although

a majority of recipients develop diabetes within the first year of transplantation, clinicians need

to also remain vigilant to monitor the risk. The organ groups at greatest risk are the lung and

multiple organ recipients and should be considered a pre-diabetic group based on underlying

chronic disease leading to end organ failure.

! 56!

7.5 Future Directions

It will be important to further explore findings generated in this study. Most importantly,

it is worth examining the effect of recurrent transplants on diabetes and the long-term impact of

diabetes on graft failure, cardiovascular events, and mortality. It is also imperative to analyze the

risk factors that are responsible for the development of diabetes in solid organ transplant

recipients. Finally, there should be guidelines for appropriate glycemic control by a

comprehensive team in children with solid organ transplantation to improve long-term

complications associated with diabetes especially in high-risk groups such as lung transplant

recipients.

!

! 57!

Acknowledgements for using ICES data

1.! This study was supported by the Institute for Clinical Evaluative Sciences (ICES) Western site. ICES is funded by an annual grant from the Ontario Ministry of Health and Long-Term Care (MOHLTC). Core funding for ICES Western is provided by the Academic Medical Organization of Southwestern Ontario (AMOSO), the Schulich School of Medicine and Dentistry (SSMD), Western University, and the Lawson Health Research Institute (LHRI). The research was conducted by members of the ICES Kidney, Dialysis and Transplantation team, at the ICES Western facility, who are supported by a grant from the Canadian Institutes of Health Research (CIHR). The opinions, results and conclusions are those of the authors and are independent from the funding sources. No endorsement by ICES, AMOSO, SSMD, LHRI, CIHR, or the MOHLTC is intended or should be inferred.

2.! Parts of this material are based on data and information compiled and provided by CIHI. However, the analyses, conclusions, opinions and statements expressed herein are those of the author, and not necessarily those of CIHI.

! 58!

References

1.! Tufton!N,!Ahmad!S,!Rolfe!C,!Rajkariar!R,!Byrne!C,!Chowdhury!TA.!NewFonset!diabetes!after! renal! transplantation.! Diabetic! medicine! :! a! journal! of! the! British! Diabetic! Association.!2014;31(11):1284F92.!2.! http://www.phacFaspc.gc.ca/cdFmc/publications/diabetesFdiabete/factsFfiguresFfaitsFchiffresF2011/highlightsFsaillantsFeng.php!F!chp1!2011![!3.! Shivaswamy! V,! Boerner! B,! Larsen! J.! PostFTransplant! Diabetes! Mellitus:! Causes,!Treatment,!and!Impact!on!Outcomes.!Endocrine!reviews.!2016;37(1):37F61.!4.! Palepu!S,!Prasad!GV.!NewFonset!diabetes!mellitus!after!kidney!transplantation:!Current!status!and!future!directions.!World!journal!of!diabetes.!2015;6(3):445F55.!5.! Garro!R,!Warshaw!B,!Felner!E.!NewFonset!diabetes!after!kidney!transplant! in!children.!Pediatric!nephrology!(Berlin,!Germany).!2015;30(3):405F16.!6.! Sharif!A,!Hecking!M,!de!Vries!AP,!Porrini! E,!Hornum!M,!RasoulFRockenschaub!S,! et! al.!Proceedings!from!an!international!consensus!meeting!on!posttransplantation!diabetes!mellitus:!recommendations!and!future!directions.!American!journal!of!transplantation!:!official!journal!of!the! American! Society! of! Transplantation! and! the! American! Society! of! Transplant! Surgeons.!2014;14(9):1992F2000.!7.! Baid! S,! Cosimi! AB,! Farrell!ML,! Schoenfeld!DA,! Feng! S,! Chung! RT,! et! al.! Posttransplant!diabetes!mellitus!in!liver!transplant!recipients:!risk!factors,!temporal!relationship!with!hepatitis!C!virus!allograft!hepatitis,!and!impact!on!mortality.!Transplantation.!2001;72(6):1066F72.!8.! Ye!X,!Kuo!HT,!Sampaio!MS,!Jiang!Y,!Bunnapradist!S.!Risk!factors!for!development!of!newFonset! diabetes! mellitus! after! transplant! in! adult! lung! transplant! recipients.! Clinical!transplantation.!2011;25(6):885F91.!9.! AlFUzri! A,! Stablein! DM,! R! AC.! Posttransplant! diabetes! mellitus! in! pediatric! renal!transplant! recipients:! a! report! of! the!North!American!Pediatric! Renal! Transplant! Cooperative!Study!(NAPRTCS).!Transplantation.!2001;72(6):1020F4.!10.! Prokai!A,! Fekete!A,! Kis! E,! Reusz!GS,! Sallay!P,! Korner!A,! et! al.! PostFtransplant!diabetes!mellitus!in!children!following!renal!transplantation.!Pediatric!transplantation.!2008;12(6):643F9.!11.! Greig! F,! Rapaport! R,! Klein! G,! Akler! G,! Annunziato! R,!Miloh! T,! et! al.! Characteristics! of!diabetes!after!pediatric!liver!transplant.!Pediatric!transplantation.!2013;17(1):27F33.!12.! Hathout! E,! Alonso! E,! Anand! R,! Martz! K,! Imseis! E,! Johnston! J,! et! al.! PostFtransplant!diabetes!mellitus! in! pediatric! liver! transplantation.! Pediatric! transplantation.! 2009;13(5):599F605.!13.! Wagner! K,! Webber! SA,! Kurland! G,! Boyle! GJ,! Miller! SA,! Cipriani! L,! et! al.! NewFonset!diabetes! mellitus! in! pediatric! thoracic! organ! recipients! receiving! tacrolimusFbased!immunosuppression.!The!Journal!of!heart!and!lung!transplantation!:!the!official!publication!of!the!International!Society!for!Heart!Transplantation.!1997;16(3):275F82.!14.! Pham! PT,! Pham! PM,! Pham! SV,! Pham! PA,! Pham! PC.! New! onset! diabetes! after!transplantation!(NODAT):!an!overview.!Diabetes,!metabolic!syndrome!and!obesity!:!targets!and!therapy.!2011;4:175F86.!15.! van! Hooff! JP,! Christiaans! MH,! van! Duijnhoven! EM.! Evaluating! mechanisms! of! postFtransplant! diabetes!mellitus.! Nephrology,! dialysis,! transplantation! :! official! publication! of! the!

! 59!

European! Dialysis! and! Transplant! Association! F! European! Renal! Association.! 2004;19! Suppl!6:vi8Fvi12.!16.! Sharif! A,! Baboolal! K.! Risk! factors! for! newFonset! diabetes! after! kidney! transplantation.!Nature!reviews!Nephrology.!2010;6(7):415F23.!17.! Kasiske! BL,! Snyder! JJ,! Gilbertson! D,! Matas! AJ.! Diabetes! mellitus! after! kidney!transplantation!in!the!United!States.!American!journal!of!transplantation!:!official!journal!of!the!American! Society! of! Transplantation! and! the! American! Society! of! Transplant! Surgeons.!2003;3(2):178F85.!18.! Paolillo! JA,! Boyle!GJ,! Law! YM,!Miller! SA,! Lawrence! K,!Wagner! K,! et! al.! Posttransplant!diabetes! mellitus! in! pediatric! thoracic! organ! recipients! receiving! tacrolimusFbased!immunosuppression.!Transplantation.!2001;71(2):252F6.!19.! BelleFvan!Meerkerk!G,!van!de!Graaf!EA,!KwakkelFvan!Erp! JM,!van!Kessel!DA,!Lammers!JW,! Biesma! DH,! et! al.! Diabetes! before! and! after! lung! transplantation! in! patients! with! cystic!fibrosis!and!other!lung!diseases.!Diabetic!medicine!:!a!journal!of!the!British!Diabetic!Association.!2012;29(8):e159F62.!20.! BendelFStenzel! MR,! Kashtan! CE,! Sutherland! DE,! Chavers! BM.! Simultaneous! pancreasFkidney! transplant! in! two! children! with! hemolyticFuremic! syndrome.! Pediatric! nephrology!(Berlin,!Germany).!1997;11(4):485F7.!21.! de!Mattos!AM,!Olyaei!AJ,!Prather! JC,!Golconda!MS,!Barry! JM,!Norman!DJ.!AutosomalFdominant! polycystic! kidney! disease! as! a! risk! factor! for! diabetes! mellitus! following! renal!transplantation.!Kidney!international.!2005;67(2):714F20.!22.! Hjelmesaeth!J,!Hartmann!A,!Kofstad!J,!Stenstrom!J,!Leivestad!T,!Egeland!T,!et!al.!Glucose!intolerance! after! renal! transplantation! depends! upon! prednisolone! dose! and! recipient! age.!Transplantation.!1997;64(7):979F83.!23.! Hur!KY,!Kim!MS,!Kim!YS,!Kang!ES,!Nam!JH,!Kim!SH,!et!al.!Risk!factors!associated!with!the!onset! and! progression! of! posttransplantation! diabetes! in! renal! allograft! recipients.! Diabetes!care.!2007;30(3):609F15.!24.! Kang!ES,!Kim!MS,!Kim!YS,!Hur!KY,!Han!SJ,!Nam!CM,!et!al.!A!variant!of!the!transcription!factor! 7Flike! 2! (TCF7L2)! gene! and! the! risk! of! posttransplantation! diabetes! mellitus! in! renal!allograft!recipients.!Diabetes!care.!2008;31(1):63F8.!25.! Ghisdal! L,! Baron! C,! Le! Meur! Y,! Lionet! A,! Halimi! JM,! Rerolle! JP,! et! al.! TCF7L2!polymorphism! associates! with! newFonset! diabetes! after! transplantation.! Journal! of! the!American!Society!of!Nephrology!:!JASN.!2009;20(11):2459F67.!26.! Kang!ES,!Kim!MS,!Kim!CH,!Nam!CM,!Han!SJ,!Hur!KY,!et!al.!Association!of!common!type!2!diabetes! risk! gene! variants! and! posttransplantation! diabetes! mellitus! in! renal! allograft!recipients!in!Korea.!Transplantation.!2009;88(5):693F8.!27.! Bamoulid! J,! Courivaud!C,!Deschamps!M,!Mercier! P,! Ferrand!C,! Penfornis!A,! et! al.! ILF6!promoter! polymorphism! F174! is! associated! with! newFonset! diabetes! after! transplantation.!Journal!of!the!American!Society!of!Nephrology!:!JASN.!2006;17(8):2333F40.!28.! Bloom!RD,!Crutchlow!MF.!TransplantFassociated!hyperglycemia.!Transplantation!reviews!(Orlando,!Fla).!2008;22(1):39F51.!29.! Knight! SR,! Morris! PJ.! Steroid! avoidance! or! withdrawal! after! renal! transplantation!increases! the! risk! of! acute! rejection! but! decreases! cardiovascular! risk.! A! metaFanalysis.!Transplantation.!2010;89(1):1F14.!

! 60!

30.! Sarwal! MM,! Ettenger! RB,! Dharnidharka! V,! Benfield! M,! Mathias! R,! Portale! A,! et! al.!Complete! steroid! avoidance! is! effective! and! safe! in! children! with! renal! transplants:! a!multicenter! randomized! trial!with! threeFyear! followFup.!American! journal!of! transplantation! :!official! journal! of! the! American! Society! of! Transplantation! and! the! American! Society! of!Transplant!Surgeons.!2012;12(10):2719F29.!31.! Jindal! RM,! Sidner! RA,! Milgrom! ML.! PostFtransplant! diabetes! mellitus.! The! role! of!immunosuppression.!Drug!safety.!1997;16(4):242F57.!32.! Filler! G,! Neuschulz! I,! Vollmer! I,! Amendt! P,! Hocher! B.! Tacrolimus! reversibly! reduces!insulin!secretion!in!paediatric!renal!transplant!recipients.!Nephrology,!dialysis,!transplantation!:!official! publication! of! the! European! Dialysis! and! Transplant! Association! F! European! Renal!Association.!2000;15(6):867F71.!33.! Vincenti!F,!Friman!S,!Scheuermann!E,!Rostaing!L,!Jenssen!T,!Campistol!JM,!et!al.!Results!of! an! international,! randomized! trial! comparing! glucose! metabolism! disorders! and! outcome!with!cyclosporine!versus!tacrolimus.!American!journal!of!transplantation!:!official!journal!of!the!American! Society! of! Transplantation! and! the! American! Society! of! Transplant! Surgeons.!2007;7(6):1506F14.!34.! RodriguezFRodriguez!AE,!Trinanes!J,!VelazquezFGarcia!S,!Porrini!E,!Vega!Prieto!MJ,!Diez!Fuentes!ML,! et! al.! The!higher!diabetogenic! risk!of! tacrolimus!depends!on!preFexisting! insulin!resistance.!A!study!in!obese!and!lean!Zucker!rats.!American!journal!of!transplantation!:!official!journal! of! the! American! Society! of! Transplantation! and! the! American! Society! of! Transplant!Surgeons.!2013;13(7):1665F75.!35.! Johnston! O,! Rose! CL,! Webster! AC,! Gill! JS.! Sirolimus! is! associated! with! newFonset!diabetes!in!kidney!transplant!recipients.!Journal!of!the!American!Society!of!Nephrology!:!JASN.!2008;19(7):1411F8.!36.! Teutonico!A,!Schena!PF,!Di!Paolo!S.!Glucose!metabolism!in!renal!transplant!recipients:!effect!of!calcineurin!inhibitor!withdrawal!and!conversion!to!sirolimus.!Journal!of!the!American!Society!of!Nephrology!:!JASN.!2005;16(10):3128F35.!37.! Crutchlow!MF,! Bloom!RD.! TransplantFassociated! hyperglycemia:! a! new! look! at! an! old!problem.!Clinical!journal!of!the!American!Society!of!Nephrology!:!CJASN.!2007;2(2):343F55.!38.! Aasebo!W,!Midtvedt! K,! Valderhaug! TG,! Leivestad! T,! Hartmann! A,! Reisaeter! AV,! et! al.!Impaired!glucose!homeostasis!in!renal!transplant!recipients!receiving!basiliximab.!Nephrology,!dialysis,! transplantation! :! official! publication! of! the! European! Dialysis! and! Transplant!Association!F!European!Renal!Association.!2010;25(4):1289F93.!39.! Ogden! CL,! Carroll! MD,! Curtin! LR,! McDowell! MA,! Tabak! CJ,! Flegal! KM.! Prevalence! of!overweight!and!obesity!in!the!United!States,!1999F2004.!Jama.!2006;295(13):1549F55.!40.! Hannon! TS,! Rao! G,! Arslanian! SA.! Childhood! obesity! and! type! 2! diabetes! mellitus.!Pediatrics.!2005;116(2):473F80.!41.! Van!Laecke!S,!Nagler!EV,!Taes!Y,!Van!Biesen!W,!Peeters!P,!Vanholder!R.!The!effect!of!magnesium! supplements! on! early! postFtransplantation! glucose! metabolism:! a! randomized!controlled!trial.!Transpl!Int.!2014;27(9):895F902.!42.! Cheungpasitporn! W,! Thongprayoon! C,! Harindhanavudhi! T,! Edmonds! PJ,! Erickson! SB.!Hypomagnesemia! linked! to! newFonset! diabetes! mellitus! after! kidney! transplantation:! a!systematic!review!and!metaFanalysis.!Endocrine!research.!2016:1F5.!

! 61!

43.! Dzurik! R,! Stefikova! K,! Spustova! V,! Fetkovska! N.! The! role! of! magnesium! deficiency! in!insulin!resistance:!an!in!vitro!study.!J!Hypertens!Suppl.!1991;9(6):S312F3.!44.! Grafton! G,! Baxter! MA.! The! role! of! magnesium! in! diabetes! mellitus.! A! possible!mechanism! for! the! development! of! diabetic! complications.! J! Diabetes! Complications.!1992;6(2):143F9.!45.! Barbagallo! M,! Dominguez! LJ.! Magnesium! metabolism! in! type! 2! diabetes! mellitus,!metabolic! syndrome! and! insulin! resistance.! Archives! of! biochemistry! and! biophysics.!2007;458(1):40F7.!46.! Filler! G,! Amendt! P,! Kohnert! KD,! Devaux! S,! Ehrich! JH.! Glucose! tolerance! and! insulin!secretion! in! children! before! and! during! recombinant! growth! hormone! treatment.! Hormone!research.!1998;50(1):32F7.!47.! Shishido!S,!Sato!H,!Asanuma!H,!Shindo!M,!Hataya!H,!Ishikura!K,!et!al.!Unexpectedly!high!prevalence! of! pretransplant! abnormal! glucose! tolerance! in! pediatric! kidney! transplant!recipients.!Pediatric!transplantation.!2006;10(1):67F73.!48.! Executive! summary:! Standards! of! medical! care! in! diabetesFF2013.! Diabetes! care.!2013;36!Suppl!1:S4F10.!49.! Pham!PT,!Pham!PC,!Lipshutz!GS,!Wilkinson!AH.!New!onset!diabetes!mellitus!after!solid!organ!transplantation.!Endocrinology!and!metabolism!clinics!of!North!America.!2007;36(4):873F90;!vii.!50.! Hux! JE,! Ivis!F,!Flintoft!V,!Bica!A.!Diabetes! in!Ontario:!determination!of!prevalence!and!incidence!using!a!validated!administrative!data!algorithm.!Diabetes!care.!2002;25(3):512F6.!51.! Guttmann! A,! Nakhla! M,! Henderson! M,! To! T,! Daneman! D,! CauchFDudek! K,! et! al.!Validation!of!a!health!administrative!data!algorithm!for!assessing!the!epidemiology!of!diabetes!in!Canadian!children.!Pediatric!diabetes.!2010;11(2):122F8.!52.! Dabelea!D,!DeGroat!J,!Sorrelman!C,!Glass!M,!Percy!CA,!Avery!C,!et!al.!Diabetes!in!Navajo!youth:! prevalence,! incidence,! and! clinical! characteristics:! the! SEARCH! for! Diabetes! in! Youth!Study.!Diabetes!care.!2009;32!Suppl!2:S141F7.!53.! Hsia! Y,! Neubert! AC,! Rani! F,! Viner! RM,! Hindmarsh! PC,! Wong! IC.! An! increase! in! the!prevalence!of!type!1!and!2!diabetes!in!children!and!adolescents:!results!from!prescription!data!from!a!UK!general!practice!database.!British!journal!of!clinical!pharmacology.!2009;67(2):242F9.!54.! Cox! ER,! Halloran! DR,! Homan! SM,!Welliver! S,! Mager! DE.! Trends! in! the! prevalence! of!chronic!medication!use!in!children:!2002F2005.!Pediatrics.!2008;122(5):e1053F61.!55.! Blanchard! JF,! Dean! H,! Anderson! K,! Wajda! A,! Ludwig! S,! Depew! N.! Incidence! and!prevalence!of!diabetes! in!children!aged!0F14!years! in!Manitoba,!Canada,!1985F1993.!Diabetes!care.!1997;20(4):512F5.!56.! Hjelmesaeth! J,! Hartmann! A,! Leivestad! T,! Holdaas! H,! Sagedal! S,! Olstad! M,! et! al.! The!impact! of! earlyFdiagnosed! newFonset! postFtransplantation! diabetes! mellitus! on! survival! and!major!cardiac!events.!Kidney!international.!2006;69(3):588F95.!57.! Koshy!SM,!Guttmann!A,!Hebert!D,!Parkes!RK,!Logan!AG.! Incidence!and!risk! factors! for!cardiovascular!events!and!death!in!pediatric!renal!transplant!patients:!a!single!center!longFterm!outcome!study.!Pediatric!transplantation.!2009;13(8):1027F33.!58.! Sumrani! NB,! Delaney! V,! Ding! ZK,! Davis! R,! Daskalakis! P,! Friedman! EA,! et! al.! Diabetes!mellitus! after! renal! transplantation! in! the! cyclosporine! eraFFan! analysis! of! risk! factors.!Transplantation.!1991;51(2):343F7.!

! 62!

59.! Gourishankar!S,!Jhangri!GS,!Tonelli!M,!Wales!LH,!Cockfield!SM.!Development!of!diabetes!mellitus! following! kidney! transplantation:! a! Canadian! experience.! American! journal! of!transplantation! :!official! journal!of! the!American!Society!of!Transplantation!and!the!American!Society!of!Transplant!Surgeons.!2004;4(11):1876F82.!60.! Benchimol! EI,! Smeeth! L,! Guttmann! A,! Harron! K,! Moher! D,! Petersen! I,! et! al.! The!REporting!of!studies!Conducted!using!Observational!RoutinelyFcollected!health!Data!(RECORD)!statement.!PLoS!medicine.!2015;12(10):e1001885.!61.! http://www.statcan.gc.ca/pub/75F202Fx/2010000/analysisFanalysesFeng.htm!2015![!62.! Regelmann!MO,! Goldis!M,! Arnon! R.! NewFonset! diabetes!mellitus! after! pediatric! liver!transplantation.!Pediatric!transplantation.!2015;19(5):452F9.!63.! Kuo!HT,! Lau! C,! Sampaio!MS,! Bunnapradist! S.! Pretransplant! risk! factors! for! newFonset!diabetes!mellitus!after!transplant!in!pediatric!liver!transplant!recipients.!Liver!transplantation!:!official! publication! of! the! American! Association! for! the! Study! of! Liver! Diseases! and! the!International!Liver!Transplantation!Society.!2010;16(11):1249F56.!64.! Rhodes! ET,! Laffel! LM,! Gonzalez! TV,! Ludwig! DS.! Accuracy! of! administrative! coding! for!type!2!diabetes!in!children,!adolescents,!and!young!adults.!Diabetes!care.!2007;30(1):141F3.!65.! Dart! AB,! Martens! PJ,! Sellers! EA,! Brownell! MD,! Rigatto! C,! Dean! HJ.! Validation! of! a!pediatric! diabetes! case! definition! using! administrative! health! data! in! manitoba,! Canada.!Diabetes!care.!2011;34(4):898F903.!66.! Kasiske!BL,!Zeier!MG,!Chapman!JR,!Craig!JC,!Ekberg!H,!Garvey!CA,!et!al.!KDIGO!clinical!practice!guideline!for!the!care!of!kidney!transplant!recipients:!a!summary.!Kidney!international.!2010;77(4):299F311.!!!!!!!!!!!!!!!!!!!!!!

63

The RECORD statement – checklist of items, extended from the STROBE statement, that should be reported in observational studies using routinely collected health data. Item

No. STROBE items Location in

manuscript where items are reported

RECORD items Location in manuscript where items are reported

Title and abstract 1 (a) Indicate the study’s design

with a commonly used term in the title or the abstract (b) Provide in the abstract an informative and balanced summary of what was done and what was found

RECORD 1.1: The type of data used should be specified in the title or abstract. When possible, the name of the databases used should be included. RECORD 1.2: If applicable, the geographic region and timeframe within which the study took place should be reported in the title or abstract. RECORD 1.3: If linkage between databases was conducted for the study, this should be clearly stated in the title or abstract.

Page ii (Abstract) Page ii (Abstract)

Introduction Background rationale

2 Explain the scientific background and rationale for the investigation being reported

Page 1-17

Objectives 3 State specific objectives, including any prespecified hypotheses

Page 18

Methods Study Design 4 Present key elements of study

design early in the paper Page 19

Setting 5 Describe the setting, locations, and relevant dates, including periods of recruitment, exposure, follow-up, and data collection

Page 19-20

64

Participants 6 (a) Cohort study - Give the eligibility criteria, and the sources and methods of selection of participants. Describe methods of follow-up Case-control study - Give the eligibility criteria, and the sources and methods of case ascertainment and control selection. Give the rationale for the choice of cases and controls Cross-sectional study - Give the eligibility criteria, and the sources and methods of selection of participants (b) Cohort study - For matched studies, give matching criteria and number of exposed and unexposed Case-control study - For matched studies, give matching criteria and the number of controls per case

RECORD 6.1: The methods of study population selection (such as codes or algorithms used to identify subjects) should be listed in detail. If this is not possible, an explanation should be provided. RECORD 6.2: Any validation studies of the codes or algorithms used to select the population should be referenced. If validation was conducted for this study and not published elsewhere, detailed methods and results should be provided. RECORD 6.3: If the study involved linkage of databases, consider use of a flow diagram or other graphical display to demonstrate the data linkage process, including the number of individuals with linked data at each stage.

Page 22 Page 22

Variables 7 Clearly define all outcomes, exposures, predictors, potential confounders, and effect modifiers. Give diagnostic criteria, if applicable.

RECORD 7.1: A complete list of codes and algorithms used to classify exposures, outcomes, confounders, and effect modifiers should be provided. If these cannot be reported, an explanation should be provided.

Page 21, 24-25

Data sources/ measurement

8 For each variable of interest, give sources of data and details of methods of assessment (measurement). Describe comparability of assessment methods if there is more than one group

Page 20-21 & 24-25

65

Bias 9 Describe any efforts to address potential sources of bias

Page 20

Study size 10 Explain how the study size was arrived at

Page 27-28

Quantitative variables

11 Explain how quantitative variables were handled in the analyses. If applicable, describe which groupings were chosen, and why

Page 24-25

Statistical methods

12 (a) Describe all statistical methods, including those used to control for confounding (b) Describe any methods used to examine subgroups and interactions (c) Explain how missing data were addressed (d) Cohort study - If applicable, explain how loss to follow-up was addressed Case-control study - If applicable, explain how matching of cases and controls was addressed Cross-sectional study - If applicable, describe analytical methods taking account of sampling strategy (e) Describe any sensitivity analyses

Page 25-27

Data access and cleaning methods

.. RECORD 12.1: Authors should describe the extent to which the investigators had access to the database population used to create the study population.

NA

66

RECORD 12.2: Authors should provide information on the data cleaning methods used in the study.

Linkage .. RECORD 12.3: State whether the study included person-level, institutional-level, or other data linkage across two or more databases. The methods of linkage and methods of linkage quality evaluation should be provided.

Page 20-21

Results Participants 13 (a) Report the numbers of

individuals at each stage of the study (e.g., numbers potentially eligible, examined for eligibility, confirmed eligible, included in the study, completing follow-up, and analysed) (b) Give reasons for non-participation at each stage. (c) Consider use of a flow diagram

RECORD 13.1: Describe in detail the selection of the persons included in the study (i.e., study population selection) including filtering based on data quality, data availability and linkage. The selection of included persons can be described in the text and/or by means of the study flow diagram.

Page 24 and 29

Descriptive data 14 (a) Give characteristics of study participants (e.g., demographic, clinical, social) and information on exposures and potential confounders (b) Indicate the number of participants with missing data for each variable of interest (c) Cohort study - summarise follow-up time (e.g., average and total amount)

Page 30-31

Outcome data 15 Cohort study - Report numbers of outcome events or summary measures over time

Page 32-43

67

Case-control study - Report numbers in each exposure category, or summary measures of exposure Cross-sectional study - Report numbers of outcome events or summary measures

Main results 16 (a) Give unadjusted estimates and, if applicable, confounder-adjusted estimates and their precision (e.g., 95% confidence interval). Make clear which confounders were adjusted for and why they were included (b) Report category boundaries when continuous variables were categorized (c) If relevant, consider translating estimates of relative risk into absolute risk for a meaningful time period

Page 32-43

Other analyses 17 Report other analyses done—e.g., analyses of subgroups and interactions, and sensitivity analyses

Page 45-48

Discussion Key results 18 Summarise key results with

reference to study objectives Page 49

Limitations 19 Discuss limitations of the study, taking into account sources of potential bias or imprecision. Discuss both direction and magnitude of any potential bias

RECORD 19.1: Discuss the implications of using data that were not created or collected to answer the specific research question(s). Include discussion of misclassification bias, unmeasured confounding, missing data, and changing eligibility over time, as they pertain to the study being reported.

Page 53

68

Interpretation 20 Give a cautious overall interpretation of results considering objectives, limitations, multiplicity of analyses, results from similar studies, and other relevant evidence

Page 54-55

Generalisability 21 Discuss the generalisability (external validity) of the study results

Page 53

Other Information Funding 22 Give the source of funding and

the role of the funders for the present study and, if applicable, for the original study on which the present article is based

NA

Accessibility of protocol, raw data, and programming code

.. RECORD 22.1: Authors should provide information on how to access any supplemental information such as the study protocol, raw data, or programming code.

NA

*Reference: Benchimol EI, Smeeth L, Guttmann A, Harron K, Moher D, Petersen I, Sørensen HT, von Elm E, Langan SM, the RECORD Working Committee. The REporting of studies Conducted using Observational Routinely-collected health Data (RECORD) Statement. PLoS Medicine 2015; in press. *Checklist is protected under Creative Commons Attribution (CC BY) license.