The 22-item Positive Health questionnaire (PH22) is reliable and applicable among different (Dutch) hospital populations: A test-retest study design

BACKGROUND: There is a growing adoption in (healthcare) organizations of Positive Health, a new theoretical concept of health moving from a disease-oriented perspective toward a more holistic understanding of health. Alongside this, there is a growing need to make the impact of working according to this concept measurable. The validated 22-item Positive Health questionnaire (PH22) has good potential for this purpose. However, information about the test-retest reliability of the PH22 is lacking. Also, its performance in patient populations is unknown. OBJECTIVES: In addition to the previous validation study in a general population, this study aimed, in line with international guidelines, to evaluate the (test-retest) reliability of the PH22, and its performance among two different hospital populations. Additionally, its discriminative ability was examined by assessing PH22 scores across subgroups defined by educational level and level of (chronic) disease. METHODS: Participants were recruited from a general hospital population and a lifestyle outpatient population with overweight. Using a test-retest design, the measurement properties: internal consistency, agreement (intraclass correlation coefficient (ICC), measurement error (ME), smallest detectable change (SDC)) and model fit indices were examined using Confirmatory Factor Analysis (CFI, RMSEA, SRMR). RESULTS: Included were 360 and 139 patients, respectively. Of whom 247 and 121 also completed the retest. Internal consistency of the PH22 was good, except for the dimension Future perspective (Cronbach's α < 0.7). Agreement was good (ICC > 0.7), except for the dimension Future perspective among the general hospital population (ICC = 0.63, ME = 2.8 (7.3%), SDC = 7.8 (26%)). Overall, the PH22 model fit was acceptable, but for the Lifestyle outpatient population only the RSMEA index was sufficient. As expected on forehand, the PH22 scores were higher among those higher educated and lower for those with a chronic disease or higher healthcare use. CONCLUSION: The PH22 appears to be a reliable and applicable instrument for measuring Positive Health among general hospital populations. Caution should be taken among more specific (patients) groups, especially for the dimension Future perspective. Future research should focus on the clinical relevance and interpretation of different PH22 scores.