Article
Development and validation of the PROPERmed instrument to identify older patients in general practice at risk of worsening of quality of life: a meta-analysis of individual participant data (IPD-MA)
Search Medline for
Authors
-
Ana Isabel González-González
- Johann Wolfgang Goethe University, Institute of General Practice, Deutschland
-
Andreas Meid
- University Hospital, Department of Clinical Pharmacology and Pharmacoepidemiology, Deutschland
-
Truci Sophia Nguyen
- Johann Wolfgang Goethe University, Institute of General Practice, Deutschland
-
Jeanet Wilhelmina Blom
- Leiden University Medical Center, Department of Public Health and Primary Care, Leiden, Niederlande
-
Marjan van den Akker
- Johann Wolfgang Goethe University, Institute of General Practice, Deutschland; School of CAPHRI, Department of Family Medicine, Maastricht University, Niederlande
-
Petra Elders
- Vrije Universiteit, Amsterdam Public Health Research Institute, Department of General Practice and Elderly Care Medicine, Amsterdam UMC, Amsterdam, Niederlande
-
Ulrich Thiem
- University of Hamburg, Albertinen-Diakoniewerk, Hamburg, Deutschland
-
Daniela Küllenberg de Gaudry
- Faculty of Medicine – University of Freiburg, Institute for Evidence in Medicine (for Cochrane Germany Foundation), Medical Center – University of Freiburg, Freiburg, Deutschland
-
Kym Snell
- Keele University, Centre for Prognosis Research, Research Institute for Primary Care & Health Sciences, Großbritannien
-
Rafael Perera
- Oxford University, Nuffield Department of Primary Care, Oxford, Großbritannien
-
Karin Swart
- Vrije Universiteit, Amsterdam Public Health Research Institute, Department of General Practice and Elderly Care Medicine, Amsterdam UMC, Amsterdam, Niederlande
-
Henrik Rudolf
- Ruhr University, AMIB, Deutschland
-
Hans-Joachim Trampisch
- Ruhr University, AMIB, Deutschland
-
Joerg Meerpohl
- Faculty of Medicine – University of Freiburg, Institute for Evidence in Medicine (for Cochrane Germany Foundation), Medical Center – University of Freiburg, Freiburg, Deutschland
-
Benno Flaig
- Johann Wolfgang Goethe University, Institute of General Practice, Deutschland
-
Ghainsom Kom
- Techniker Krankenkasse (TK), Deutschland
-
Walter Emil Haefeli
- University Hospital, Department of Clinical Pharmacology and Pharmacoepidemiology, Deutschland
-
Paul Glasziou
- Bond University, Centre for Research in Evidence-Based Practice, Australien
-
Ferdinand Gerlach
- Johann Wolfgang Goethe University, Institute of General Practice, Deutschland
-
Christiane Muth
- Johann Wolfgang Goethe University, Institute of General Practice, Deutschland
| Published: | February 12, 2020 |
|---|
Outline
Text
Background/research question: Multimorbidity and polypharmacy increase the risk for potentially inappropriate medications (PIMs), underuse and worsening of health-related quality of life (wHRQoL). In heterogeneous general practice populations, it would be helpful to identify those patients at high risk of wHRQoL, since they may benefit most from resource-intensive interventions. Therefore, we developed and internal validated a prognostic model to predict wHRQoL in a six-month period in older patients in general practice with at least one chronic condition and one chronic prescription.
Methods: We harmonized IPD from five cluster-randomized trials conducted in the Netherlands and Germany. wHRQoL was defined as a decrease of EQ-5D-3L index score ≥5% from baseline to 6-month follow-up. Prognostic variables included socio-demographics, lifestyle, morbidity, medication, functional status and well-being (HRQoL, pain, depressive symptoms). Model development used logistic regression with a stratified-intercept to account for between-study heterogeneity. Prognostic variables were selected in complete cases and refitted in multiply imputed data to obtain the final model. Internal validation applied bootstrapping to assess reproducibility and internal-external cross-validation (IECV) was used to evaluate generalisability.
Results: The complete-case population consisted of 3,582 patients. In 1,046 (29%) patients, HRQoL worsened after 6-months. Predictors in the final model were related to single conditions (i.e. coronary heart disease), certain medication (i.e. drugs for acid-related disorders), PIMs (i.e. systemic corticosteroids in COPD), underuse (e.g. ACE inhibitors in heart failure), functional status and well-being, with most prognostic relevance attributable to baseline HRQoL and functional status (82%). Internal validation showed good discrimination (C-statistic: 0.71; 95%CI 0.69 to 0.72) and calibration (calibration slope: 0.97, 95%CI 0.54 to 1.39). In IECV, the final model performed well (pooled C-statistic: 0.69, 95%CI 0.68 to 0.71; calibration-in-the-large: 0, 95%CI -0.12 to 0.13).
Conclusion: This first IPD-based prognostic model for wHRQoL in older patients with chronic conditions and medication in general practice performed well in discrimination, calibration, and generalisability and may thus help to identify high-risk patients. After external validation, implementation may be supported by the relative importance of HRQoL and functional status.
Competing interests: This work was supported by the German Innovation Funds according to § 92a (2) Volume V of the Social Insurance Code (§ 92a Abs. 2, SGB V - Fünftes Buch Sozialgesetzbuch), grant number: 01VSF16018.The funder had no role in developing the protocol for this review.
Authors do not have any conflict of interest to declare.
