1. A NATION-WIDE PROJECT FOR THE REVISION OF THE BELGIAN NURSING MINIMUM DATASET: FROM CONCEPT TO IMPLEMENTATION Walter SERMEUS, PhD, RN Katholieke Universiteit LEUVEN BELGIUM 22nd MIC conference Brussels November 25, 2004

2. Co-authors & acknowledgement Research team –Leuven University: W. Sermeus, L. Delesie, K. Vanden Heede, D. Michiels –Liège University: P. Gillet, J. Codognoto, O. Thonon, C. Van Boven Commissioned by the Ministry of Public Health, Belgium

3. Nursing Minimum Data Set Belgium Compulsory from 1988 All Belgian acute Hospitals Content: –Patient demographics –23 nursing interventions –Nurse staffing data (FTE nurses / qualification level) Sample: 15 days / 3 months 19 Million data recorded since 1988 Largest Nursing Dataset in the world

4. General rules for revision 2002 - 2005 International language Integrated in hospital discharge dataset Basic + specific set per care programme: cardiology, oncology, paediatrics, geriatrics, intensive care, chronic care, one-day clinic Multi-purpose: reimbursement, nurse staffing, appropriateness evaluation, quality management Minimal High user involvement

5. Phase I: Development of conceptual framework 2002/6 – 2002/10 Methods: –Literature review –Secondary data-analysis Choice for NIC-framework: –Nursing intervention Classification (NIC): 433 interventions in 26 classes –Comprehensive, research-based, standardised, international accepted, available in French and Dutch

6. Phase II: Language development 2002/11 – 2003/9 Methods: –Panels of clinical experts (N= 75) –Six care programmes: cardiology, geriatrics, oncology, rehabilitation, paediatrics, intensive care –Pre-test in 15 different hospitals Results: –Alpha version: 105 interventions –Definitions & scoring manual

7. Phase III: Pilot testing and tool validation 2003/10 – 2004/12 Methods: –Data collection (dec.2003 – march 2004) –30 days of data collection –66 hospitals, 158 nursing wards, 95000 patient records –NMDS-I, NMDS-II, DRGs, Financial data (Finhosta), Pharmaceutical data Results: –Number of interventions/pat/day: Med=14 (1- 43) –Time (N= 3504: 42 hospitals, 81) wards: Med= 4’ (IQR=3’- 7’ ) –Interrater reliability (9 cases, 66 raters): Above 70% for 80% of the interventions

8. Validity testing Criterion-related validity: –comparing NMDS-II (49) with NMDS-I (19) –21146 patient records –Correlations from 0,88 (tube feeding) to 0,16 (emotional support) Construct validity: –Principal component analysis (CATPCA) –Intra-class and interclass analysis –Latent variables: nursing intensity, care/cure Face validity: –Expert panels (2004/11) are validating final NMDS-II

9. Interim results (2004/11) –Core dataset: Comparing between care programs Between 25 + 39 variables –Supplementary datasets: Comparing within care programs Paediatrics: 32 + 9 Intensive care: 31 + 12 Geriatrics: 36 + 5 Rehabilitation & Chronic care: 37 + 5 Cardiac care: 34 + 8 Oncology: 39 + 4 One Day clinic: 25 + 5

10. Phase IV: From data to applications 2004/9 – 2005/9 Appropriateness Evaluation protocol (AEP) –Nursing care explains 80% of stay in hospital –In B-NMDS: 8 nursing interventions support AEP Nurse Staffing –Validating patients’ nursing care profiles in relation to nurse staffing ratios (N=100, 30 raters, planning march 2005) –Selection of key-interventions for nurse staffing Reimbursement / funding –Nurse staffing & nursing care profile per DRG –2004/10 – 2005/6 Quality Management –Nurse staffing sensitive patient outcomes (Needleman et.al. 2002) –Nurse intervention – patient problem relations (in 2005)

11. AEP – NMDS II EXPLAINING APPROPRIATE STAY Source: Gillet et.al., 2004

12. Linking NMDS and DRGs N days-of-stay in HDDS Nursing profile / day-of-stay * Data: DRG 166, Coronary Artery Bypass Graft (CABG) without cardiac catherisation Y2000 Belgian data, 5575 admissions, 74925 inpatient days, 2670 linked NMDS- days

13. Per Hospital ? For DRG166 Comparing 2 hospitals A - B Hospital A: Median LOS=9 days Hospital B: Median LOS=12 days Number of admission: 584/y Number of inpatient days: 6377 N in NMDS: 399 (6,2%) Number of admission: 379/y Number of inpatient days: 5009 N in NMDS: 322 (6,4%)

14. NMDS – nursing profile for hospital B N in NMDS : 322 N in NMDS: 399 NMDS – nursing profile for hospital A

15. Relationship NMDS-profile and nurse staffing ratios 1 ---------------------------------------------------------------- 7 1/10 1/9 1/8 1/7 1/6 1/5 ¼ 1/3 ½ 1/1 Nurses/patient/day (selfcare Highly dependent) Study March, 2005 Delphi on NMDS-profiles (k=100, r=30) Analysing actual staffing ratios in hospitals

16. Phase V: From application to implementation 2005/9 – 2006/6 Education Legislation Informatics Control programs Audit procedures …

17. Added value of NMDS-II ? 1. Updating nursing language according to evolutions in clinical practice (interventions, education, …) 2. Differentiation in core and supplementary datasets : –Core datasets: comparing nursing care in different care programs within the organisation –Supplementary datasets: comparing nursing care in different hospitals within a care program 3. NIC-framework: international comparisons 1.International benchmarking 2.Standardisation of nursing language 3.Integrating NIC / NMDS in electronic patient records

18. Added value of NMDS-II ? 4. Integration of NMDS other datasets 1.Simplification in data transmission 2.Enhanced control systems 5. Linking with hospital discharge dataset 1.Nursing care profiles per DRG 2.New applications for funding and reimbursement 6. New applications 1.E.g. Appropriate Evaluation Protocols 2.Quality Monitoring 7. NMDS-II is using open framework of data collection 1.Easy update according to new technology, interventions,…

19. Added value of NMDS-II ? 8. NMDS-II offers a nurse staffing application 1.Benchmarking nurse staffing between hospitals 2.Comparing actual and required nurse staffing levels 9. NMDS-II will complement the Finhosta dataset 1.Dynamic (NMDS-II) versus static (Finhosta) nurse staffing data 10. NMDS-II is congruent to NMDS-I 1.Continuity in the data 2.+ Wider in description and application 3.+ More in-depth, refined measurement