Definition of SGI/SSMI-categories

using NEMS and SAS

Kommission Patientenklassifzierung der SGI - KPK

Commission pour la classification des patients - CCP

Berne, March 14th 2006

A simple scoring system:General requirements

For general use in all Swiss ICU‘s

Simple

Objective and reproducible

Validated (i.e. measures what it is intended to)

Scoring the patient (i.e. assigning a patient to a SGI/SSMI-category) should be simple, uniform and objective

Scoring• Is required to document the patient process

(as required by KAI, KWFB and paritätische Kommission)• Serves as basis for the reimbursement system (Tarmed)• May be used to calculate indicators for quality management• Might serve as basis for an ICU-specific module in Swiss-DRG• May have some limited use as a tool in human resource

management

A simple scoring system:General requirements

Scoring the patient (i.e. assigning a patient to a SGI/SSMI-category) should be simple, uniform and objective

Scoring• Is required to document the patient process

(as required by KAI, KWFB and paritätische Kommission)• Serves as basis for the reimbursement system (Tarmed)• May be used to calculate indicators for quality management• Might serve as basis for an ICU-specific module in Swiss-DRG• May have some limited use as a tool in human resource

management

Taylor C J et al. Taylor C J et al. WorkloadWorkload managementmanagement in an in an emergencyemergency departmentdepartment::countingcounting thethe uncountableuncountable and and predictingpredicting thethe unpredictableunpredictable..J J AccidAccid EmergEmerg MedMed 1997, 14:88-91. 1997, 14:88-91.

Hughes M. Hughes M. NursingNursing workloadworkload: an : an unquantifiableunquantifiable entityentity..J J NursNurs Manag 1999, 7:317-22. Manag 1999, 7:317-22.

A simple scoring system:General requirements

SGI/SSMI-categories: Acutal definitions

Der Pflegepersonalbedarf auf Intensivstationen, Schriftenreihe SKI Bd. 41 1989, p.72ff(see also: Critical Care Medicine, JAMA 1983, 250:798-804)

< 20< 20

20 20 –– 29 29

30 30 –– 39 39

> 40> 40

ZeitbedarfZeitbedarf(Min/h)(Min/h)

< 8< 8

8 8 –– 12 12

12 12 –– 16 16

> 16> 16

ZeitbedarfZeitbedarf(h/Tag)(h/Tag)

33

22

1B1B

1A1A

KategorieKategorie

1/31/3

2/32/3

3/33/3

4/34/3

PflegepersonalPflegepersonalpro Schichtpro Schicht

619874

6571027

621882

17263

Kat. 1BErste 2 Std.Jede weitere 8 Std.

619340

657493

621348

17216

Kat. 3Erste 2 Std.Jede weitere 8 Std.

619608

657760

621615

17228

Kat. 2Erste 2 Std.Jede weitere 8 Std.

6191141

6571295

6211150

172239

Kat. 1AErste 2 Std.Jede weitere 8 Std.

NichtärztlicheLeistungNeo-IPS

NichtärztlicheLeistung

Verbr.-IPS

NichtärztlicheLeistung

IPS

ÄrztlicheLeistung

N.B.: nichtärztliche Leistung Kardio-/angiolgische Überwachungsstation: erste 2 Std. 82TP, dann32TP pro Std., maximal 14 mal

Kategoriewechsel jeweils bei Schichtwechsel

Tarmed 1.2 (ab 01.01.05)

Intensivpflegestationen (IPS)

Situation 01.01.06HU. Rothen

Swiss ICU‘s with SGI/SSMI-acknoweldgement

New calibration of SGI/SSMI-categories:participating units

Number of shifts per participating ICUs

5.3%11

4.5%12

9.6%137.6%

14

10.6%16

7.8%21

5.9%22

15.4%23

15.6%24 2.3%

252.9%26

12.5%27

Distribution of assigned SGI-categories (0 = 1A, 1 = 1B, 2 = 2, 3 = 3)

18.8%0

41.9%1

31.2%2

8.1%3

Distribution of SGI/SSMI-categoriesin participating units

1A1A1B1B

22

33

Distribution of assigned SGI-categories for each individual ICU11

0

1

2

3

12

0

12

3

130

1 2

3

14

0

1

2

3160

1

2

3

210

1

2

3

220

1

2

3

23

0

1

2

3

24

0

123

25

0

1

2

3

26

0

1

2

3

27

0

1

2

3

Distribution of SGI/SSMI-categoriesin participating units

0 10 20 30 40 50 60NEMS (poins)

0

1000

2000

3000

4000

5000

Count

0.0

0.1

0.2

0.3

0.4

Proportion per Bar

Distribution of NEMSin participating units

(points)

Distribution of RASS

4.1%-5

9.3%-4

6.5%-34.8%

-210.7%

-1

42.7%0

10.8%1

8.0%2

2.3%30.7%4

Distribution of SAS

7.6%1

5.6%2

7.8%3

61.7%4

14.3%5

2.3%60.7%7

Distribution of SAS/RASSin participating units

n = 9‘154 n = 2‘515

3.0% 3.0%

Definition of SGI/SSMI-categoriesusing NEMS and SAS

241-320 0-160161-240>320 Min.6 (LEP)

321B1AModel

0-20

0-12

0-12

0-13

0-13

0-12

0-10

0-16

21-30

21-30

19-33

19-33

27-33

22-32

26-32

21-30

13-20

13-20

14-18

14-18

13-26

11-21

17-25

>30Euricus-1

>30 Pt.5_S (NEMS)

>30 Pt.5 (NEMS)

>33 Pt.4_S (NEMS)

>33 Pt.4 (NEMS)

>33 Pt.3 (NEMS)

>32 Pt.2 (NEMS)

>32 Pt.1 (NEMS)

SGI/SSMI-category

models 4_S and 5_S: category increased by 1 level if SAS > 5

Definition of SGI/SSMI-categoriesusing NEMS and SAS

Distribution of SGI-categories (original and 6 models. 0: 1A, 1: 1B, 2: 2, 3: 3)

20.6%0

43.4%1

27.8%2

8.2%3

24.6%019.4%1

43.0%2

13.0%3

24.6%0

30.2%1

42.0%2

3.2%3

23.3%020.6%1

50.0%2

6.1%3

23.3%033.4%1

37.2%2

6.1%3

26.2%0

30.0%1

37.8%2

6.1%3

52.6%0

21.7%1 18.5%2

7.2%3

Assigned („old“) Model 1 Model 2

Model 3 Model 4 Model 5 Model LEPn = 4‘675

Distribution of SGI-categories (original and 6 models. 0: 1A, 1: 1B, 2: 2, 3: 3)

18.8%041.9%1

31.2%2

8.1%3

23.6%018.9%1

41.9%2

15.7%3

23.6%029.4%1

43.4%2

3.6%3

22.3%020.1%1

50.8%2

6.8%3

22.3%033.5%1

37.3%2

6.8%3

25.1%0

30.3%1

37.8%2

6.8%3

52.6%0

21.7%1 18.5%2

7.2%3

0 10 20 30 40 50 60NEMS (points)

0

1000

2000

3000

4000

Count

0.0

0.1

0.2

0.3

Proportion per Bar

3210

NEWCAT_5

SGI/SSMI-categoriesusing model 5

-4 -2 0 2 4assigned category - modelled category 0

0

2000

4000

6000

Count

Change in SGI/SSMI-category„old“ vs. model 5 including SAS

Note: negative number = modelled (“new”) category is lower

assigned category – modelled catergory

Distribution of assigned SGI-categories for each individual ICU11 12 13 14

16 21 22 23

24 25 26 27

Distribution of SGI/SSMI-categories„old“

Distribution of SGI-categories (model 5, incl. effect of SAS) for each individual ICU11 12 13 14

16 21 22 23

24 25 26 27

Distribution of SGI/SSMI-categoriesmodel 5 including SAS

Patient Scoring:An uniform system for Switzerland?

Common standard for Switzerland Common documentation of patient process and performance

(KAI, KWFB, paritätische Kommission für WB in IP) Common language in MDSi

(Quality managment, Benchmarking, etc.) Common base for reimbursement

(SLK/TarMed/SwissDRG?)

The actual used scoring system (SGI/SSMI-categories) is not veryplausible, there is thus a risk of low credibility in the future

Units using LEP or PRN would prefer not to add a further, newscoring system

Patient scoring (categorization)using NEMS and SAS/RASS

NEMS and SAS can easily be collected with minimal amount of expenses

NEMS and SAS form a common base for all users

NEMS is well established in the general medical literature

NEMS is validated as a tool to analyse the patient process in critical care

NEMS may be used to calculate indicators for quality management

NEMS allows for a direct, easy documentation of a few corecompetencies in critical care

NEMS (and SAS) might serve as basis for an ICU-specific module inSwiss-DRG

NEMS (and SAS) may have some limited use as tool in human resourcemanagement

Patient scoring (categorization)using NEMS and SAS/RASS

Due to the small number of items included, there is limited resolution

NEMS is not a tool for detailed assessment of nursing care

If used without any further context, neither NEMS nor SAS will allow toestimate quality of care

NEMS and SAS are not included in TarMed

NEMS is in general used only once per 24 hours

SAS/RASS have only recently been introduced

The point of view of the patient is virtually missing

Patient scoring (categorization)using model 5_S

Aims at an overall similar distribution of SGI-categories as acutallyused („conventional“ categorization)

SAS is included based on a proposal by nurses

Units using RASS may easily transform their data to SAS-equivalents

Model 5_S shows plausible distribution of categories, both in non-university and in university ICUs

Patient scoring (categorization)using model 5_S

Possibly might result in a small increase in the relative number ofshifts with SGI-category 1A, and on the other hand a small decreasein SGI-category 3

The share of category 1A/1B is high in comparison to EURICUS-1(but is low in comparison to the LEP-model)

The model needs further evaluation Use in paediatric/neonatology ICU

Large-scale application

Explore the future use of this model in intermediate care units

Explore the use of this model (or NEMS) for Swiss-DRG

The KPK proposesto calculate SGI/SSMI-categories based on NEMS and SAS:• NEMS (nine equivalents of nursing manpower use score) contains

nine elements, representing core competencies of critical care• SAS (sedation agitation scale) or RASS (Richmond agitation-

sedation scale) is used to assess the patient‘s mental state

Categories ar defined as followsCategory NEMS (points)

• 1A > 30• 1B 21 – 30• 2 13 – 20• 3 0 – 12• If SAS > level 5, a given category is increased by 1 step.

Note that category 1A can not be increased further. The MDSi-tool calculates the SGI/SSMI-categories based on

this new algorithm since 01.01.06

Patient scoring (categorization)using NEMS and SAS/RASS

NEMS: Nine equivalents of nursing manpower use score• Miranda D R at al. Nine equivalents of nursing manpower use score (NEMS).

Intensive Care Med 1997, 23:760-765.• Iapichino G et al. Description of trends in the course of illness of critically ill patients. Markers of

intensive care organization and performance.Intensive Care Med 2002, 28:985-9.

• Hartmann B et al. Vollautomatische Kalkulation des Pflegeaufwandes auf einer OperativenIntensivstation. Ein Vergleich zwischen TISS-28 und NEMS.Intensivmed 2004, 41:94-8.

SAS: Sedation agitation scale• Riker RR et al. Prospective evaluation of the Sedation-Agitation Scale for adult critically ill patients.

Crit Care Med 1999, 27:1325-9• Riker RR et al. Validating the Sedation-Agitation Scale with the Bispectral Index and Visual Analog Scale

in adult ICU patients after cardiac surgery.Intensive Care Med 2001, 27:853-8

• Brandl KM et al. Confirming the reliability of the sedation-agitation scale administered by ICU nurseswithout experience in its use.Pharmacotherapy 2001, 21:431-6.

Patient scoring (categorization)using NEMS and SAS