1 23

International Journal of ClinicalPharmacyInternational Journal of ClinicalPharmacy and Pharmaceutical Care ISSN 2210-7703 Int J Clin PharmDOI 10.1007/s11096-013-9864-y

Adherence to medication for chronicdisorders during pregnancy: results from amultinational study

Angela Lupattelli, Olav Spigset & HedvigNordeng

1 23

Your article is protected by copyright and

all rights are held exclusively by Koninklijke

Nederlandse Maatschappij ter bevordering

der Pharmacie. This e-offprint is for personal

use only and shall not be self-archived

in electronic repositories. If you wish to

self-archive your article, please use the

accepted manuscript version for posting on

your own website. You may further deposit

the accepted manuscript version in any

repository, provided it is only made publicly

available 12 months after official publication

or later and provided acknowledgement is

given to the original source of publication

and a link is inserted to the published article

on Springer's website. The link must be

accompanied by the following text: "The final

publication is available at link.springer.com”.

RESEARCH ARTICLE

Adherence to medication for chronic disorders during pregnancy:results from a multinational study

Angela Lupattelli • Olav Spigset • Hedvig Nordeng

Received: 27 May 2013 / Accepted: 1 October 2013

� Koninklijke Nederlandse Maatschappij ter bevordering der Pharmacie 2013

Abstract Background For a variety of chronic disorders,

low medication adherence during pregnancy may jeopar-

dize maternal as well as foetal health. Little is known about

how closely pregnant women follow their chronic phar-

macotherapy regimens. Objective To explore the level of

adherence to medication for a variety of chronic disorders,

namely cardiovascular, rheumatic and bowel disorders,

diabetes and epilepsy, during pregnancy and to identify

determinants of low adherence during pregnancy. Setting

This multinational, cross-sectional, internet-based study

was undertaken in 18 countries in Europe, North America

and Australia. Data originating from some South American

countries were also collected. Methods The study period

lasted from 1-October-2011 to 29-February-2012. By using

an anonymous on-line questionnaire we collected infor-

mation about maternal demographics, chronic disorders

and related medication use during pregnancy, and women’s

pregnancy-specific beliefs about medication. Main out-

come measure Adherence to medication during pregnancy

via the 8-item Morisky Medication Adherence Scale

(MMAS-8). Results A total of 210 pregnant women

reported chronic medication use during pregnancy and

filled in the MMAS-8. Overall, 36.2 % had low medication

adherence. On the basis of the MMAS-8, the rates of low

adherence were 55.6 % for medication for rheumatic dis-

orders, 40.0 % for epilepsy, 36.1 % for bowel disorders,

32.9 % for cardiovascular disorders, and 17.1 % for dia-

betes. A lack of folic acid use, having previous children,

and individual pregnancy-specific beliefs about medication

were significant determinants of low medication adherence

during pregnancy. Conclusion Many pregnant women had

low adherence to their chronic pharmacotherapy regimens

during pregnancy. Women’s beliefs about medication were

a central factor determining low adherence.

Keywords Adherence � International study �Medication � Pregnancy

Impact on practice

• Many pregnant women present low adherence to their

chronic pharmacotherapy regimens during pregnancy.

For some disorders, maternal-fetal health may thereby

be jeopardized.

• Healthcare providers should be more aware of the role

of women’s beliefs about medication in relation to

medication adherence.

• Interventions proven to enhance medication adherence

in the general population should also be implemented in

maternity care.

Electronic supplementary material The online version of thisarticle (doi:10.1007/s11096-013-9864-y) contains supplementarymaterial, which is available to authorized users.

A. Lupattelli (&) � H. Nordeng

Department of Pharmacy, School of Pharmacy, University of

Oslo, Blindern, PO Box 1068, 0316 Oslo, Norway

e-mail: angela.lupattelli@farmasi.uio.no

O. Spigset

Department of Clinical Pharmacology, St Olav’s University

Hospital, Trondheim, Norway

O. Spigset

Department of Laboratory Medicine, Children’s and Women’s

Health, Norwegian University of Science and Technology,

Trondheim, Norway

H. Nordeng

Division of Mental Health, Norwegian Institute of Public Health,

Oslo, Norway

123

Int J Clin Pharm

DOI 10.1007/s11096-013-9864-y

Author's personal copy

Introduction

The use of medications for chronic disorders is estimated to

occur in 26 % of all pregnancies [1]. It is well established that

conditions such as diabetes, hypothyroidism, epilepsy or

chronic hypertension may lead to adverse pregnancy outcomes

if such health conditions are suboptimally treated [2–5]. Hence,

appropriate clinical management of women’s chronic diseases

during pregnancy is central for maternal-foetal health.

Poor adherence to chronic therapies in the general

population is a well-known public health concern [6].

Unfortunately, little is known about medication adherence

in pregnant subjects [7]. Fear of teratogenic drug effects

may possibly result in even lower adherence to prescribed

treatments in pregnancy [8]. On the other hand, pregnant

women may be more attentive to follow recommended

therapies for the benefit of their unborn child.

Previous research indicates that 40.9 % of women

adhered to their chronic treatment during pregnancy [1],

whereas higher rates (59.7–67.0 %) were observed among

pregnant women with ulcerative colitis or Crohn’s disease

[9, 10]. Yet, the degree of medication adherence during

pregnancy and risk factors for non-adherence still needs to

be elucidated for most chronic diseases during pregnancy.

Identification and understanding of such determinants may

assist in the development of strategies for the prevention of

non-adherence in pregnancy and help healthcare personnel

to identify non-adherent patients.

Aim of the study

The aim of this study was to explore the level of adherence

to medication for a variety of chronic disorders, namely

cardiovascular, rheumatic and bowel disorders, diabetes

and epilepsy, during pregnancy and to identify determi-

nants of low adherence during pregnancy.

Ethical approval

Informed consent was given by participants by ticking the

answer ‘‘yes’’ to the question ‘‘Are you willing to partici-

pate in the study?’’ The study was approved by the

Regional Ethics Committee in South-East Norway. All data

were reported, handled and stored anonymously.

Method

Study design and data collection

This is a sub-study of a multinational, cross-sectional,

internet-based study performed in 18 countries in Western,

Northern and Eastern Europe, North America, and Aus-

tralia. Data originating from some South American coun-

tries were also collected. Pregnant women at any

gestational week were included in the analysis. An anon-

ymous on-line questionnaire (http://www.questback.com)

was utilized for data collection, accessible for a period of

2 months in each participating country between 1-Oct-

2011 and 29-Feb-2012.

The questionnaire was open to the public via utilization of

banners (invitations to participate in the study) on national

websites and/or social networks commonly visited and con-

sulted by pregnant women and/or new mothers (cf.

e-Table 1). The questionnaire was first developed in Nor-

wegian and English and then translated into the other relevant

languages. A pilot study in four countries (n = 47) elicited no

major changes. Collected data were scrutinized for the pre-

sence of potential duplicates (based on reported country of

residency, socio-demographics, date and time of question-

naire completion) but none were identified. Data selection to

achieve the final study sample is outlined in Fig. 1.

Chronic medication users during pregnancy

Participants were first presented with a list of chronic disor-

ders, among them cardiovascular (i.e. hypertension, high

cholesterol, heart diseases) and rheumatic disorders (i.e.

rheumatoid arthritis, psoriatic arthritis), diabetes (type I and

II) and epilepsy. An open-ended option was also available,

where any other condition not previously listed could be

specified. Women were then questioned about medication use

for each individual disorder as a free-text entry. All recorded

medications were coded into the corresponding Anatomical

Therapeutic Chemical (ATC) codes in accordance with the

World Health Organization (WHO) ATC index [11].

Pregnant women who indicated that they suffered from a

cardiovascular or rheumatic disorder, diabetes or epilepsy,

or were specifying bowel disorders (i.e. ulcerative colitis,

Crohn’s disease, inflammatory bowel syndrome or irritable

bowel syndrome) or other cardiovascular disorders (i.e.

deep vein thrombosis and thrombophilia) under the open-

ended option were selected for the data analysis. Women

reporting the use of medications for any of these disorders

were classified as chronic medication users (Fig. 1). We

used the Food and Drug Administration (FDA) pregnancy

risk classification system to categorize all medications used

during pregnancy according to risk to the fetus [12].

In situations of multiple medication use, we assigned the

category of the medication with highest pregnancy risk.

Medication adherence during pregnancy

Adherence was measured by using the 8-item Morisky

Medication Adherence Scale (MMAS-8), a structured, self-

Int J Clin Pharm

123

Author's personal copy

reported medication adherence measure with satisfactory

internal consistency (Cronbach’s alpha reliability of 0.83)

[13, 14]. For each self-reported chronic disorder, a disorder-

specific MMAS-8 was available (e.g. women reporting to

suffer from diabetes were presented with a specific MMAS-

8 adapted to antidiabetic medications). Validated translated

versions of the original English MMAS-8 were available for

eight languages other than English. For the remaining six

languages, translation into the relevant language and back-

translation to English was done by two independent native

speakers and/or translators. Professor Donald E. Morisky

approved the construct validity of all translated and/or

adapted items of the MMAS-8 (Professor Donald E. Mori-

sky, personal communication).

For each disorder-specific MMAS-8, the sum score

(ranging from 0 to 8) was calculated and then trichotom-

ized into low (sum score \6), medium (sum score 6 or 7)

and high (sum score 8) adherence [13]. Imputed values

were generated when respondents completed at least six of

the eight items on the MMAS-8 using the estimation-

maximization algorithm [15]. Values were imputed for

2.9 % of the study population.

Maternal characteristics

Maternal characteristics included gestational week, previ-

ous children, marital status, folic acid use before and/or

during pregnancy, unplanned pregnancy, country of resi-

dency, age, employment status, educational level and

mother tongue. Life-style characteristics included smoking

status before and during pregnancy and alcohol consump-

tion after awareness of pregnancy.

We compared socio-demographic and life-style charac-

teristics of our study population on an individual country level

with those of the general birthing population in the country.

Reports from National Statistics Bureaus or previous national

studies were utilized for this purpose (cf. e-Table 2).

Beliefs about medications

Three statements were used to explore the women’s beliefs

about medication during pregnancy: (1) ‘‘I have a higher

threshold for using medicines when I am pregnant than

when I am not pregnant’’; (2) ‘‘Even though I am ill and

could have taken medicines, it is better for the fetus that I

refrain from using them’’; (3) ‘‘Pregnant women should

preferably use herbal remedies than conventional medi-

cines’’. Study participants could strongly agree, agree,

disagree, strongly disagree, or be uncertain about each of

these statements. Responses to the statements were

trichotomized into ‘‘agree’’, ‘‘disagree’’, or ‘‘uncertain’’.

Statistical analysis

Descriptive statistics were utilized as appropriate. Pearson

Chi square or Fisher’s exact tests were used to compare the

maternal characteristics in users and non-users of relevant

chronic medication during pregnancy.

Fig. 1 Participant flow-chart to

achieve final analysis sample.

*The examined chronic

disorders include:

cardiovascular, rheumatic and

bowel disorders, diabetes and

epilepsy. �Relevant medications

include medications for

treatment of cardiovascular,

rheumatic or bowel disorders,

diabetes or epilepsy

Int J Clin Pharm

123

Author's personal copy

Table 1 Maternal characteristics according to medication use status for the examined chronic disorders during pregnancy* (n = 315)

Maternal characteristics Examined chronic disorders p value Group 1

vs. Group 2Group 1

Medication use

(n = 210)

Group 2

No medication use

(n = 105)

n (%) n (%)

Region of residency

Western Europe� 52 (24.8) 25 (23.8) 0.540

Northern Europe� 86 (41.0) 40 (38.1)

Eastern Europe§ 41 (19.5) 28 (26.7)

North America** 21 (10.0) 10 (9.5)

South America�� 4 (1.9) –

Australia 6 (2.9) 2 (1.9)

Maternal age (years)

B20 3 (1.4) 5 (4.8) 0.251

21-30 112 (53.3) 60 (57.1)

31-40 93 (44.3) 39 (37.1)

C41 2 (1.0) 1 (1.0)

Time of gestation

1st trimester 57 (27.3) 26 (24.8) 0.884

2nd trimester 89 (42.6) 47 (44.8)

3rd trimester 63 (30.1) 32 (30.5)

Previous children

No 93 (44.3) 57 (54.3) 0.094

Yes 117 (55.7) 48 (45.7)

Marital status

Married/cohabiting 199 (94.8) 95 (90.5) 0.151

Single/divorced/others 11 (5.2) 10 (9.5)

Folic acid use��

Yes 194 (94.2) 93 (89.4) 0.132

No 12 (5.8) 11 (10.6)

Working status

Employed, but not as health care professional 120 (57.1) 51 (48.6) 0.246

Health care professional 33 (15.7) 21 (20.0)

Student 16 (7.6) 11 (10.5)

Housewife 21 (10.0) 6 (5.7)

Job seeker 6 (2.9) 7 (6.7)

Other than above 14 (6.7) 9 (8.6)

Highest educational level

High school 59 (28.1) 27 (25.7) 0.824

Higher than high school 13 (6.2) 5 (4.8)

Lower than high school 116 (55.2) 59 (56.2)

Others, unspecified 22 (10.5) 14 (13.3)

Alcohol use after awareness of pregnancy

No 184 (87.6) 96 (91.4) 0.310

Yes 26 (12.4) 9 (8.6)

Smoking before pregnancy

No 136 (65.1) 65 (61.9) 0.581

Yes 73 (34.9) 40 (38.1)

Int J Clin Pharm

123

Author's personal copy

Univariate and multivariate logistic regression was used

to explore determinants of low adherence during pregnancy.

A p value of\0.05 was considered statistically significant.

Data are presented as crude (OR) and adjusted odds ratios

(aOR) with 95 % confidence intervals (CI). The multivariate

model was built after fitting the univariate logistic regression

model for all explanatory variables. Adjustment was done for

relevant confounding variables (age, time of gestation,

having previous children, educational level). The Hosmer

and Lemeshow test was used to assess goodness of fit of the

final multivariate model [16]. The MMAS-8 internal con-

sistency was assessed via reliability analysis (Cronbach’s

alpha) [17]. All statistical analyses were performed by using

the Statistical Package for the Social Sciences (SPSS) ver-

sion 20.0 (IBM� SPSS� Statistics).

Results

Population characteristics

A total of 5,166 pregnant women accessed the on-line

questionnaire and 5,095 (98.6 %) completed it. Women

with no eligible country of residency were excluded,

leaving 5,089 participants. Overall, the birthing population

in each participating country was reflected quite well by the

sample with respect to age and smoking habits (e-Table 2).

However, on average, the women in the study had higher

education and were slightly more often primiparous than

the general birthing population in each country.

Of the 5,089 participants, 315 pregnant women self-

reported to suffer from at least one of the chronic disorders

examined and completed the MMAS-8 (Fig. 1). Of these,

210 (66.7 %) reported treatment with relevant medications

during pregnancy. Maternal characteristics based on medi-

cation use are shown in Table 1. Most participants (272/

315; 86.3 %) were residing in Europe at the time of com-

pletion of the questionnaire.

Medication use

The self-reported prevalence of the examined disorders and

related medications used are outlined in e-Table 3. A total

of 13 women reported concomitant chronic disorders dur-

ing pregnancy. These co-morbidities were rheumatic/bowel

disorders (n = 4), cardiovascular disorders/diabetes

(n = 4), cardiovascular/rheumatic disorders (n = 2), car-

diovascular/bowel disorders (n = 2), and epilepsy/diabetes

(n = 1).

Medications belonging to the FDA-assigned pregnancy

categories D or X were used by 90 women (42.9 %), while

the remaining 120 (57.1 %) used medications of category

A, B or C. The use of medications under category D/X

occurred most frequently among women with rheumatic

Table 1 continued

Maternal characteristics Examined chronic disorders p value Group 1

vs. Group 2Group

1Medication

use(n = 210)

Group 2 No

medication

use(n = 105)

n (%) n (%)

Smoking during pregnancy

No 181 (86.6) 94 (89.5) 0.473

Yes 28 (13.4) 11 (10.5)

Planned pregnancy

Yes 195 (92.5) 92 (87.6) 0.124

No 15 (7.1) 13 (12.4)

Immigrant status§§

No 197 (93.8) 105 (100.0) 0.006

Yes 12 (6.2) –

Numbers may not add up to total due to missing values* Examined chronic disorders include: cardiovascular, rheumatic or bowel disorders, diabetes and epilepsy� Western Europe includes Austria, France, Italy, Switzerland, The Netherlands and United Kingdom� Northern Europe includes Finland, Iceland, Norway and Sweden§ Eastern Europe includes Croatia, Poland, Russia, Serbia and Slovenia** North America includes USA and Canada�� South America includes Uruguay�� Indicates folic acid use before and/or during pregnancy§§ Women having the first language different from the official main language in the country of residency

Int J Clin Pharm

123

Author's personal copy

disorders (66.7 %), epilepsy (60.0 %) and bowel disorders

(52.8 %), whereas the use was lower for cardiovascular

disorders (35.4 %) and very low for diabetes (2.9 %).

Medication adherence

Overall, 76 women (36.2 %) had low adherence to their

chronic medication regimens during pregnancy. The level of

medication adherence during pregnancy by type of chronic

disorder is outlined in Table 2. All MMAS-8 scores had

satisfactory internal consistency (Cronbach’s alpha [ 0.7)

(Table 2). The rates of low adherence ranged from 55.6 %

for treatment of rheumatic disorders to 17.1 % for diabetes.

In a corollary analysis, we compared women who reported

treatment with medications for symptomatic cardiovascular

disorders such as venous thrombosis (i.e. those treated with

antithrombotic agents) with women who reported to be

medicated for asymptomatic disorders such as hypertension

(i.e. those treated with beta-blockers and calcium channel

blockers). We found the first group’s adherence to their

therapeutic regimens (high: 45.5 %; medium: 36.4 %; low:

18.2 %) was significantly higher (Chi square test,

p = 0.031) than those in the second group (high: 22.4 %;

medium: 34.7 %; low: 42.9 %).

Determinants of low adherence

The level of agreement with the three pregnancy-specific

beliefs among women with low adherence is depicted in

Fig. 2. Pregnant subjects agreeing with the statement that it

is better to abstain from using medication whilst pregnant

despite being ill, presented a significant increased likeli-

hood of low adherence (OR 2.17, 95 % CI 1.09–4.34)

compared to those women who disagreed (p = 0.028).

Similarly, women agreeing with the statement that herbal

remedies rather than conventional medications should be

used during pregnancy, presented a significant increased

likelihood of low adherence (OR 3.74, 95 % CI 1.73–8.06)

compared to those who disagreed (p = 0.001). No statis-

tically significant difference was found between the two

groups with respect to the remaining statement.

Determinants of low medication adherence during preg-

nancy are presented in e-Table 4. In the multivariate ana-

lysis, women with previous children were more likely to

poorly adhere to chronic medications than women with no

previous children (aOR 2.04, 95 % CI 1.09–3.84). Similarly,

pregnant women who did not use folic acid before and/or

during pregnancy had a significantly increased likelihood of

low adherence (aOR 5.13, 95 % CI 1.25–21.11). Region of

residency did not significantly influence adherence. Use of

medications belonging to FDA-assigned pregnancy catego-

ries D/X did not have any significant impact on adherence

(aOR 1.10, 95 % CI 0.59–2.06).

Discussion

Medication adherence

This is the first multinational study utilizing a validated

instrument, the MMAS-8, to explore medication adherence

during pregnancy. Several findings are important for clin-

ical practice. Firstly, the extent of non-adherence in preg-

nancy was high (36.2 %). This finding aligns with previous

research observing an overall rate of non-adherence equal

to 40.9 % during pregnancy [1]. The level of medication

adherence varied substantially across the chronic disorders

examined. The highest rates of low adherence were

detected for treatment of rheumatic disorders (55.6 %) and

epilepsy (40.0 %), whereas diabetes accounted for the

lowest (17.1 %). These differences might be related to

various issues including differences between disorders in

Table 2 Level of medication adherence for the examined chronic disorders and MMAS-8 reliability (Cronbach’s alpha) (n = 210)

Examined chronic disorder No. of

subjects*

Cronbach’s

alpha

Adherence

Sum Score�Low

adherence

Medium

adherence

High

adherence

n Mean ± SD n (%) n (%) n (%)

Diabetes 35 0.76 6.92 ± 1.21 6 (17.1) 16 (45.7) 13 (37.2)

Cardiovascular disorders 82 0.76 6.40 ± 1.60 27 (32.9) 29 (35.4) 26 (31.7)

Epilepsy 25 0.77 6.21 ± 1.67 10 (40.0) 11 (44.0) 4 (16.0)

Bowel disorders 36 0.80 5.94 ± 2.31 13 (36.1) 10 (27.8) 13 (36.1)

Rheumatic disorders 45 0.71 5.09 ± 2.15 25 (55.6) 13 (28.9) 7 (15.5)

MMAS-8: 8-item Morisky Medication Adherence Scale; SD standard deviation* Number of subjects does not add up to 210 because 13 women were medication users for more than one chronic disorder.� MMAS-8 sum score can range from 0 to 8

Use of the �MMAS is protected by US copyright laws. Permission for use is required. A Licensure agreement is available from: Donald E.

Morisky, ScD, ScM, MSPH, Professor, Department of Community Health Sciences, UCLA School of Public Health, 650 Charles E. Young Drive

South, Los Angeles, CA 90095-1772, dmorisky@ucla.edu

Int J Clin Pharm

123

Author's personal copy

terms of the risk/benefit evaluations in pregnancy, organi-

zation of obstetric follow-up, fluctuations of the disorder,

and teratogenic risk of medications. For example, the high

proportion of low adherence in the treatment of bowel and

especially rheumatic disorders could be ascribed to a qui-

escent phase or amelioration of these conditions during

pregnancy [18–20]. Our estimate of low medication

adherence to treatment of bowel disorders, though slightly

higher, aligned with findings of previous research indicat-

ing non-adherence to be 35.5 and 20.0 % among patients

with ulcerative colitis and Crohn’s disease, respectively [9,

10]. In our study, inclusion of a milder condition such as

irritable bowel disorder in the bowel disorder group might

have inflated our estimate.

The high rate of low adherence for treatment of epilepsy is

also noteworthy, raising concerns about its suboptimal

treatment during pregnancy. Previous research has shown

that women with epilepsy were frequently non-compliant to

anticonvulsant medications during pregnancy (62.3 %) and

might greatly reduce or even stop their prescribed medica-

tion [21, 22]. Even though a seizure-free period may lead

pregnant women to not adhere to their therapeutic regimens,

the driving cause of low medication adherence may in fact

reside within the fear of the teratogenic risk of anticonvul-

sants. On the other hand, most women with epilepsy will be

followed by neurologists, and they should be among the most

well-informed patients regarding the risks of the underlying

illness and the necessity of the medication [23].

In our study, a clinically relevant proportion of women

(17.1 %) might be considered at risk for suboptimal blood

glucose control during pregnancy due to low adherence to

their antidiabetic regimes. High adherence to antidiabetic

medications is necessary during pregnancy in order to

achieve normoglycaemia and reduce the risk of adverse

pregnancy outcomes [4, 24].

Determinants of low adherence

We identified several risk factors for non-adherence; lack

of folic acid use and having previous children increased the

risk of low adherence during pregnancy by two-fold and

five-fold, respectively. This could reflect a past experience

with non-adherence in pregnancy without sequelae, lower

awareness regarding risks in general and/or lower will-

ingness to follow recommendations by healthcare profes-

sionals. Women’s region of residency did not significantly

influence non-adherence to chronic medication during

pregnancy, emphasising that poor adherence is a world-

wide public health concern [6]. The use of medication

under FDA pregnancy category D/X did not have any

significant impact on adherence during pregnancy. Thus,

factors other than the intrinsic medication risk may be more

important for adherence, such as e.g. the severity of illness,

the perceived need for treatment, and individual risk/ben-

efit considerations. Such an assumption is also substanti-

ated by the association between certain beliefs and low

adherence identified in the current study. The beliefs that it

is better for the fetus to abstain from using medication

whilst pregnant despite being ill and that herbal remedies

should be preferred to conventional medications during

pregnancy were both powerful determinants of low

adherence. This was also demonstrated among non-

Fig. 2 Beliefs about

medication use among pregnant

women with low adherence to

medications for their chronic

disorder. Statement 1: ‘‘I have a

higher threshold for using

medicines when I am pregnant

than when I am not pregnant’’;

Statement 2: ‘‘Even though I am

ill and could have taken

medicines, it is better for the

fetus that I refrain from using

them’’; Statement 3: ‘‘Pregnant

women should preferably use

herbal remedies than

conventional medicines’’

Int J Clin Pharm

123

Author's personal copy

pregnant subjects, where low adherence was correlated

with a patient’s concerns about the prescribed medication

and the belief that it might be harmful [25, 26]. Adequate

counselling and proper risk communication may attenuate

women’s negative beliefs about medication and heighten

medication adherence during pregnancy [27, 28], although

the benefit of interventions proven to improve medication

adherence in the general population should also be tested

among the pregnant population. Moreover, the role of

healthcare personnel counselling on adherence in preg-

nancy warrants further investigation.

Strengths and limitations

An important strength of the study is the use of a validated

questionnaire for self-reporting of medication adherence,

with satisfactory internal consistency. The simultaneous

collection of data on medication adherence for the treatment

of various chronic disorders allows for inter-disorder com-

parability. The identification of maternal characteristics and

pregnancy-specific beliefs associated with low adherence

during pregnancy enabled us to detect groups of women that

more likely need information about adequate control of

chronic disorders during pregnancy and tailored professional

counselling. By restricting our analysis to pregnant women

only, we limited the risk of recall bias. Furthermore, women

may feel more comfortable in answering sensitive questions

truthfully in an anonymous, web-based questionnaire rather

than in a face-to-face interview.

The main limitation of the study is the lack of validity of

the diagnoses. The chronic disorders examined were self-

reported by the participants and hence, dependent on the

woman’s perception of the medical condition. Information

about medication use during pregnancy was also dependent

on the accuracy of the woman’s reporting. The sample

sizes for the specific disorders were small, thus limiting the

statistical power. The questionnaire was only available

through internet websites, which did not permit calculation

of a conventional response rate. However, recent epide-

miological studies indicate reasonable validity of web-

based recruitment methods [29, 30]. Also, the penetration

rate of the internet, either in households or at work, is

relatively high among women of childbearing age [31–35].

Hence, the degree to which our findings can be extrapo-

lated to the target population is based on the representa-

tiveness of the respondents to the general birthing

populations in each country. On average, the women in the

study had higher education and were slightly more often

primiparous than the general birthing populations. Women

in need of information about medication use during preg-

nancy might have been more likely to consult internet

websites and, therefore, to participate in the present study.

Lastly, the MMAS-8 has not been validated among

pregnant subjects. The development and validation of a

medication adherence instrument specifically designed for

pregnant subjects should be definitely addressed by future

research studies.

Conclusion

A substantial proportion of women had low adherence to

their chronic pharmacotherapy regimens during pregnancy,

raising concern about suboptimal control of the underlying

maternal illness. This is especially important for treatments

where the untreated disease may be more harmful than the

prescribed medication to the unborn child. A lack of folic

acid use, having previous children and individual beliefs

about medication, were important determinants of low

adherence during pregnancy. Adequate counselling and

proper teratogenic risk communication will potentially

attenuate women’s negative beliefs about medication and

heighten medication adherence during pregnancy.

Increased awareness of women’s beliefs about medication

is needed among healthcare providers.

Acknowledgments We thank the Scientific Board of OTIS and

ENTIS, the website providers who contributed to the recruitment

phase, the national coordinators of the study (Twigg MJ, Zagorod-

nikova K, Mardby AC, Moretti ME, Drozd M, Panchaud A, Hameen-

Anttila K, Rieutord A, Gjergja Juraski R, Odalovic M, Kennedy D,

Rudolf G, Juch H, Passier JLM and Bjornsdottir I) and all partici-

pating women. We also thank Professor Donald E. Morisky for letting

us use the MMAS-8.

Funding The study has received financial support from the Nor-

wegian Research Council (Grant no. 216771/F11) and the Foundation

for Promotion of Norwegian Pharmacies and the Norwegian Phar-

maceutical Society.

Conflicts of interest The authors declare no conflicts of interest.

References

1. Sawicki E, Stewart K, Wong S, Leung L, Paul E, George J.

Medication use for chronic health conditions by pregnant women

attending an Australian maternity hospital. Aust N Z J Obstet

Gynaecol. 2011;51(4):333–8.

2. Ross DS. Hypothyroidism during pregnancy: clinical manifesta-

tions, diagnosis, and treatment. In: Basow DS, editor. Up to date.

Waltham, MA: Up To Date; 2013.

3. Jovanovic L, Knopp RH, Kim H, Cefalu WT, Zhu XD, Lee YJ,

et al. Elevated pregnancy losses at high and low extremes of

maternal glucose in early normal and diabetic pregnancy: evi-

dence for a protective adaptation in diabetes. Diabetes Care.

2005;28(5):1113–7.

4. Correa A, Gilboa SM, Besser LM, Botto LD, Moore CA, Hobbs

CA, et al. Diabetes mellitus and birth defects. Am J Obstet

Gynecol. 2008;199(3):237 e1–9.

Int J Clin Pharm

123

Author's personal copy

5. ACOG practice bulletin. Chronic hypertension in pregnancy.

ACOG Committee on practice bulletins. Obstet Gynecol. 2001;

98(1):suppl 177–185.

6. Sabate E. Adherence to long-term therapies: evidence for action:

World Health Organization; 2003.

7. Matsui D. Adherence with drug therapy in pregnancy. Obstet

Gynecol Int. 2012; 2012:Article ID 796590, 5 pages.

8. Nordeng H, Ystrøm E, Einarson A. Perception of risk regarding

the use of medications and other exposures during pregnancy. Eur

J Clin Pharmacol. 2010;66(2):207–14.

9. Julsgaard M, Norgaard M, Hvas CL, Buck D, Christensen LA.

Self-reported adherence to medical treatment prior to and during

pregnancy among women with ulcerative colitis. Inflamm Bowel

Dis. 2011;17(7):1573–80.

10. Nielsen MJ, Norgaard M, Holland-Fisher P, Christensen LA.

Self-reported antenatal adherence to medical treatment among

pregnant women with crohn’s disease. Aliment Pharmacol Ther.

2010;32(1):49–58.

11. WHO Collaborating Centre for Drugs Statistics Methodology.

ATC/ddd index 2012. [cited 2012 17 March]; Available from:

http://www.whocc.no/atc_ddd_index/.

12. Briggs GG, Freeman RK, Yaffe SJ. Drugs in pregnancy and

lactation. [S.l.]: Lippincott Williams & Wilkins; 2011.

13. Morisky DE, Ang A, Krousel-Wood M, Ward HJ. Predictive

validity of a medication adherence measure in an outpatient

setting. J Clin Hypertens (Greenwich). 2008;10(5):348–54.

14. Donald Morisky’s website. Morisky medication adherence scale.

2011 [cited 2013 23 May]; Available from: http://dmorisky.bol.

ucla.edu/MMA_scale.html.

15. Dempster AP, Laird NM, Rubin DB. Maximum likelihood from

incomplete data via the em algorithm. J R Stat Soc Series B Stat

Methodol. 1977;39(1):1–38.

16. Hosmer DW, Lemeshow S. Applied logistic regression. 2nd ed.

New York: Wiley-Interscience; 2000.

17. Cronbach LJ. Coefficient alpha and the internal structure of tests.

Psychometrika. 1951;16(3):297–334.

18. Ostensen M, Villiger PM, Forger F. Interaction of pregnancy and

autoimmune rheumatic disease. Autoimmun Rev. 2012;11(6–7):

A437–46.

19. de Man YA, Hazes JM, van der Heide H, Willemsen SP, de Groot

CJ, Steegers EA, et al. Association of higher rheumatoid arthritis

disease activity during pregnancy with lower birth weight: results

of a national prospective study. Arthr Rheum. 2009;60(11):

3196–206.

20. Agret F, Cosnes J, Hassani Z, Gornet JM, Gendre JP, Lemann M,

et al. Impact of pregnancy on the clinical activity of crohn’s

disease. Aliment Pharmacol Ther. 2005;21(5):509–13.

21. Fairgrieve SD, Jackson M, Jonas P, Walshaw D, White K,

Montgomery TL, et al. Population based, prospective study of the

care of women with epilepsy in pregnancy. BMJ.

2000;321(7262):674–5.

22. Williams J, Myson V, Steward S, Jones G, Wilson JF, Kerr MP,

et al. Self-discontinuation of antiepileptic medication in preg-

nancy: detection by hair analysis. Epilepsia. 2002;43(8):824–31.

23. Schachter SC. Management of epilepsy and pregnancy. In: Ba-

sow DS, editor. Up to date. Waltham, MA: Up To Date; 2013.

24. Kitzmiller JL, Block JM, Brown FM, Catalano PM, Conway DL,

Coustan DR, et al. Managing preexisting diabetes for pregnancy:

summary of evidence and consensus recommendations for care.

Diabetes Care. 2008;31(5):1060–79.

25. Mardby AC, Akerlind I, Jorgensen T. Beliefs about medicines

and self-reported adherence among pharmacy clients. Patient

Educ Couns. 2007;69(1–3):158–64.

26. Horne R, Weinman J. Patients’ beliefs about prescribed medi-

cines and their role in adherence to treatment in chronic physical

illness. J Psychosom Res. 1999;47(6):555–67.

27. Jasper JD, Goel R, Einarson A, Gallo M, Koren G. Effects of

framing on teratogenic risk perception in pregnant women.

Lancet. 2001;358(9289):1237–8.

28. Einarson A, Selby P, Koren G. Abrupt discontinuation of psy-

chotropic drugs during pregnancy: fear of teratogenic risk and

impact of counselling. J Psychiatry Neurosci. 2001;26(1):44–8.

29. Ekman A, Dickman P, Klint A, Weiderpass E, Litton J-E. Feasi-

bility of using web-based questionnaires in large population-based

epidemiological studies. Eur J Epidemiol. 2006;21(2):103–11.

30. van Gelder MM, Bretveld RW, Roeleveld N. Web-based ques-

tionnaires: the future in epidemiology? Am J Epidemiol. 2010;

172(11):1292–8.

31. Seybert H. Internet use in households and by individuals in 2011.

Eurostat statistics in focus 2011.

32. Internet World Stats. Usage and population statistics. 2012 [cited

2012 13 November]; Available from: http://www.internetworld

stats.com/.

33. Australian Bureau of Statistics. Household use of information

technology, Australia, 2010-11 [cited 2012 13 November]; Avail-

able from: http://www.abs.gov.au/ausstats/abs@.nsf/Latestproducts/

8146.0Main%20Features12010-11?opendocument&tabname=Sum

mary&prodno=8146.0&issue=2010-11&num=&view=.

34. Statistics Canada. Individual internet use and e-commerce. 2010

[cited 2012 13 November]; Available from: http://www.statcan.

gc.ca/daily-quotidien/111012/dq111012a-eng.htm.

35. United States Census Bureau. The 2012 statistical abstract.

Information & communications: Internet publishing and broad-

casting and internet usage. 2010 [cited 2012 13 November];

Available from: http://www.census.gov/compendia/statab/cats/infor

mation_communications/internet_publishing_and_broadcasting_

and_internet_usage.html.

Int J Clin Pharm

123

Author's personal copy