Patterns of adoption and use of a web-based decision support system for CVD risk assessment and management in primary care
3 likes540 views
The document discusses the adoption and utilization patterns of a web-based decision support system for cardiovascular disease (CVD) risk assessment in New Zealand primary care. Despite a substantial number of clinicians adopting the Predict-CVD program, significant variations in usage frequency were observed, influenced by factors such as professional registration and years since graduation. The study provides valuable insights into the barriers and facilitators affecting the implementation of decision support technologies in general practice, with implications for future healthcare interventions.
Patterns of adoption and use of a web-based decision support system for CVD risk assessment and management in primary care
- 1. Patterns of adoption and use of a web-based decision support system for CVD risk assessment and management in primary care Sue Wells, Janine Bycroft, Ai-Wei Lee, Tim Kenealy, Tania Riddell, Rod Jackson (U of A) & Paul Roseman and Kate Moodabe, ProCare Health Ltd
- 2. Overview Background-why CVD, why decision support Study Aim & design Methods Results Conclusion
- 3. NZ Cardiovascular statistics Leading cause of death & hospital admissions in NZ 9,000 deaths from CVD each year Approx 30% under the age of 70 yrs 500 deaths on the roads, 600 deaths from breast cancer large health disparities by ethnicity and deprivation
- 4. v v Identifying high risk patients
- 5. Our Approach.. i. ‘evidence in’ - to improve practice Web-based CVD risk assessment and management programme: Fast (moment of care) User Friendly Integrated with patient electronic record Guideline and evidence based Individualised recommendations for each patient
- 6. and ii. ‘evidence out’ to improve practice-relevant evidence provide clinicans with moment of care, patient-specific evidence-based advice on assessment and management CVD and Diabetes using web-based technology simultaneously generate population evidence for evaluating health needs, assessing health care improving the accuracy of risk prediction to feedback into routine practice
- 7. 2002 University of Auckland, New Zealand Guidelines Group, National Heart Foundation, clinical team from CMDHB and ProCare, IT team from Enigma Implementation: 2002 ProCare Prompt (opportunistic RA & management) 2003 CMDHB Chronic disease management module 2004 Middlemore Hospital Coronary Care Unit & Medical Outpatients PREDICT-CVD
- 8. Implementation in ProCare Enrolled pop ~660,000, ~500 GPs, 450 practice nurses integrated with MedTech, Next Gen educational seminars to doctors voluntary participation target population according to national guidelines incentive payment $900/GP after 90 patients practice IT facilitators installed software, integration mapping, safe connectivity, one-off training
- 9. How did PREDICT CVD work? In clinical setting – doctor or practice nurse sees a patient and decides to assess & manage their CVD risk Patient-practitioner interaction
- 13. Risk management: Action Plan
- 15. Risk Management: patient information
- 16. Questions who adopted PREDICT? if they adopted it – how did they use it? any significant differences between user groups? what are the most common barriers & challenges to using PREDICT? what factors enable and support practices to start or use PREDICT more? popular images of power users and non-users – true or myths?
- 17. PREDICT Health Provider Study Aim: describe the patterns of use, barriers, challenges and enablers to the use of PREDICT by primary healthcare providers 3-part qualitative/quantitative evaluation key informant interviews ProCare staff & focus groups (GPs & nurses) Questionnaire GPs/nurses at CME/CNE group meetings (May 2007) descriptive quantitative data analysis Adopters and non-adopters patterns of use over time, characteristics of users
- 18. Methods substudy-patterns of adoption and use Data collection ProCare Clinical registry+NZ Medical Council data Personal and practice names removed, NZMC/NZNC number retained PREDICT usage data Time and date of PREDICT usage NZMC or NZNC of user Data Linkage- Registry + PREDICT usage data Ethical Approval March 2007 Northern Region Ethics Committee
- 19. Results 1: adoption 533 clinicians chose to adopt PREDICT programme doctors (416) and nurses(117) 607 clinicians did not adopt doctors (289) and nurses (318) 75% who did not adopt had compatible patient management systems No differences in adoption by gender,country of medical degree More likely to have vocational registration, and be less than 20yrs since graduation incomplete practice data on 20% doctors and over 25% nurses (funding, location, size of practice)
- 21. Results: adoption 0 0
- 22. Results 2: patterns of use over time Aug 2002-Jan 2007 45,437 CVD risk assessments were conducted on 25, 705 patients 416 GPs and 117 nurses Used multiple times within one consultation Used subsequent follow-up
- 23. Results : patterns of use New guidelines published PREDICT CVD-Diabetes released ProCare implementation updated PREDICT
- 24. Results 3: frequency of use (n= 416 GPs) 31% (n=129)non-user: less than 5 patients 23% (n=95) infrequent 5-20 patients 25% (n=104) frequent user 21-89 patients 21% ( n=88) most frequent (90+ patients)
- 25. Results: frequency of use (GPs) non-user: less than 5 patients infrequent 5-20 patients compared to frequent user 21-89 patients most frequent user (90+ patients)
- 26. Results: frequency of use (GPs) Infrequent/non-users- more likely less than 10years since graduation No difference by gender, country of medical degree Older docs (graduated 30yrs+) were as likely to be in either group Frequent/most frequent-more likely to have vocational registration Unable to adjust for FTE
- 27. Proportion of CVD risk assessments conducted – graphed overtime classified into 4 types emerged Build up then decline Start high then decline Constant All then nothing Results: patterns of use of most frequent GPs (88 GPs)
- 28. 1. Build up then decline 36/88 GPs (41%)
- 29. 2. Start high and decline 29/88 GPs (33%)
- 30. 3. Fairly constant 13/88 GPs (15%)
- 31. 4. All then nothing 10/88 GPs (11%)
- 32. This Study Lifecycle of software from first implementation to obselescence 45,000+ risk assessments, ~26,000 patients 92% usage by doctors Large variation in frequency of use No differences by gender or country of medical degree Differences found by having vocation registration and year of graduation may relect time spent working in practice Data limitations
- 33. This Study Financial incentive ?faciltated uptake, little impact on usage Payment in retrospect, no feedback on progress to target Adoption responsive to annual promotional activites via cell groups Distinct patterns of use may inform interventions for uptake and to maintain sustainable use
- 34. Conclusions PREDICT CVD -1st of its kind in NZ This study -some insights into adoption & usage of decision support technology in general practice qualitative analyses and questionnaire results more in-depth understanding for future interventions to improve uptake Learnings/feedback from PREDICT-CVD applied to new module & implementation package Currently 1000+ risk assessments/month with CVD-Diabetes