1 / 14

April 12th 2013 I. Peytremann Bridevaux, MD, MPH, DSc , K. Iglesias PhD , B. Burnand, MD, MPH.

Disentangling dimension results of the Patient Assessment of Chronic Illness Care (PACIC) instrument by using published validation models and data from Swiss diabetic patients. April 12th 2013 I. Peytremann Bridevaux, MD, MPH, DSc , K. Iglesias PhD , B. Burnand, MD, MPH.

oren
Download Presentation

April 12th 2013 I. Peytremann Bridevaux, MD, MPH, DSc , K. Iglesias PhD , B. Burnand, MD, MPH.

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Disentangling dimension results of the Patient Assessment of Chronic Illness Care (PACIC) instrument by using published validation models and data from Swiss diabetic patients April 12th 2013 I. Peytremann Bridevaux, MD, MPH, DSc, K. Iglesias PhD, B. Burnand, MD, MPH. Institute of Social and PreventiveMedicine Lausanne UniversityHospital

  2. Plan • The PACIC instrument • Aim of the study • Methods • Results • Discussion and conclusion

  3. Whatis the PACIC instrument ? • Patient Assessment of Chronic Illness Care • Self-administered questionnaire • Patients’ perspective • Questionnaire developed to assess whether care for patients with chronic diseases is congruent with the Chronic Care Model • 20 items, 5-point response scale (1=never, 2=generally not, 3=sometimes, 4=most of the time, 5=always) • Dimensions and overall PACIC are scored by simple averaging of items www.improvingchronicare.org

  4. Five original dimensions • « Patient activation » (3 items) • « Delivery system design/decision support » (3 items) • « Goal setting/tailoring » (5 items) • « Problem solving/contextual counseling » (4 items) • « Follow-up/coordination » (5 items)  Theyemphasize patient-healthcare team interactions, in particular, aspects of self-management support

  5. PACIC published literature • Versions exist in several languages : English (2005), Spanish (2008), Dutch (2008), Danish (2010), German (2011), French (2011) • Described structure (dimensions) emerging from validation studies: • 5 (original) dimensions • 2 dimensions (2 different models) • 1 dimension (20-items) • 1 dimension (11-items = PACIC short form)  … lack of consensus on PACIC structure …

  6. Aim of the study • To better understand the structure of the PACIC instrument  To test all published validation models, using one single dataset and statistical tools adapted to the ordinal structure of the data

  7. Methods • Population: 406 non-institutionalized adult patients with diabetes (canton of Vaud, Switzerland) • Instrument: French version of the 20-items PACIC • Statistical analyses: • Descriptive analysis (data quality): means, % missing, floor and ceiling effects • Confirmatory factorial analysis (CFA), based on: • Polychoric correlation matrix • Likelihood estimation with a multinomial distribution for manifest variables • Pearson estimator or variance-covariance matrix • For each model: evaluation of loadings and goodness of fit

  8. Characteristics of included patients (n=406)

  9. Descriptive data qualityanalysis • Means of item responses: 1.7 - 3.7 • Range of missing values: 5.7% - 12.3% (item 5) • Floor effect: 7% - 67% (items 10,17) • Ceiling effect: 4% - 45.5% (items 5,12)

  10. PACIC structure (CFA based on polychoriccorrelationmatrix) • Loadings were relatively high. • The only model showed acceptable to good fits was the 11-items single dimension model (RMSEA<.08, SRMR <.05, CFI >.97). • This model also appeared to be the only one presenting acceptable fits when performing the two other types of CFA.

  11. Possible explanations for the lack of consensus on the PACIC structure 1) The original 5 dimensions structure was not the right one: • Glasgow suggested it in its validation paper • Published literature suggests fewer dimensions 2) Inappropriate choice of statistical tools and selection criteria may affect results: • Can retain too many dimensions • May impact the loadings of the dimensions • Considering the magnitude of loadings as selection criteria is wrong since it is not a measure of goodness of fit 3) Sample sizes maybe too small: • Rule of thumb for CFA: >10 responses/item => minimum of 200 patients in validation analysis

  12. Possible explanations for the lack of consensus on the PACIC structure 4) PACIC versions used not always similar: • Anchoring response categories: « almost never » and « almost always » versus « never » and « always » • Response categories (5- versus 11- points scale) 5) Diversity of contexts, cultures and chronic diseases considered could impact results. However, the single dimension structure was found: • In several countries: USA, Germany, The Netherlands, Spain, CH • With various chronic diseases: diabetes, cardiovascular diseases, non-specific chronic conditions

  13. Conclusion • The model considering 11 items in a single dimension appeared to best fit our data. • The lack of consensus on the PACIC structure was linked to statistical problems rather than differences in contexts, cultures, types of chronic diseases, or PACIC versions considered. • To get an overall picture of experiences of people receiving care for chronic diseases, a single score might be used. This could be done in complement to the consideration of single item results.

  14. Thankyou for your attention !

More Related