This study investigated the direct relationship between nursing home price and QOL of nursing home residents, as knowledge about this topic is currently lacking. Literature suggested that indicators such as resident characteristics (e.g.: age, education, chronical conditions) and facility characteristics (e.g.: facility size, nursing home staffing rates) are significantly associated with QOL [7, 10, 12]. The study of Shippee et al. (2015a) is one of the first studies that subdivides QOL in different domains, in order to have a more nuanced understanding of how indicators are associated with QOL. Therefore, they use a survey consisting of 52 items measuring 6 QOL domains (environment, personal attention, food, personal engagement, negative mood and positive mood). In this study, we chose the internationally validated and widely accepted InterRAI instrument as QOL measure because of two reasons. First, the InterRAI instrument supports multiple clinical and management applications for different audiences (ex.: care planning, quality measurement, patient safety assessment) [21]. Second, it has been developed in an international context and has shown to be reliable and valid in multiple languages (also in Dutch) and countries [21, 22].
The descriptive results suggested that average price levels are statistically different between public and private nursing homes. Therefore, ownership type was used as a control variable in the regression analyses. The multiple regression analyses suggested that a 10 euro increase in prices is associated with a significant decrease of 0.1 in QOL in access to services, comfort and environment, food and meals, respect, and safety and security. When splitting up the regression analyses in private and public nursing homes, results indicate a weak association between price and the QOL domains of access to services, comfort and environment, food and meals, respect, and safety and security, but only in private nursing homes. Nevertheless, based on this data we do not find evidence of an association between price and the varying domains of QOL. This result suggests that a price increase does not always lead to higher QOL, even though it is the popular belief. This result adds evidence to the study of Shippee et al. (2015a) who only investigated the association between QOL and some financial indicators. They suggest that nursing homes with a higher percentage of Medicaid-only residents (US health insurance program only for people with a limited income); compared to Medicare (US health insurance program for people older than 65 years old and younger for people with disabilities), self-pay, and privately insured residents; are associated with lower QOL in the domains of personal attention and engagement. Besides the research question, our findings also suggest that nursing homes that have more highly dependent residents are associated with higher QOL in food and meals, respect, safety and security, and staff-resident bonding (while holding the other variables constant). In Flanders, nursing homes receive a reimbursement from the Flemish government based on the dependency rate of their residents [23]. Possibly, nursing homes spend this higher reimbursement in better patient care, which might translate in higher QOL of their residents. Furthermore, our results suggest that there is no significant association between ownership type and QOL. Meanwhile, the study of Shippee et al. (2015a) indicate the opposite, more specifically that public ownership type was positively associated with the QOL dimension of environment. The different results between our study and the study of Shippee et al. (2015a) might be explained by the varying country-specific variations in ownership type.
The findings of this study should be interpreted carefully. First of all, the means of the QOL domains in the descriptive statistics are the means of ordinal variables. It only indicates if residents are on average more or less satisfied on a QOL domain or not. The number itself does not has a meaning. Second, because of the cross-sectional design of this study we could only assess associations but no causal relationships. Third, all the nursing homes are situated in Flanders. National data on the QOL domains were not available. Therefore, the results can only be generalized on a Flemish level since the sample includes 82% of all Flemish nursing homes. Fourth, the InterRAI survey was only taken from people without cognitive problems. Due to a low response level per nursing home, residents with cognitive problems were not included in the study. Therefore, the findings cannot be generalized for residents with cognitive problems. Lastly, the sample consists of aggregated information of residents in nursing homes, but not of individual data per resident. Therefore, we lose a lot of information that could make the data more sensitive and that could uncover more relationships. Additionally, we could not control for individual differences in QOL in the same nursing home such as: health status, socio-demographic characteristics, educational level, whether residents have Alzheimer’s disease or chronic conditions, which are predictive for a number of QOL domains [24]. However, we tried to compensate this loss by including the dependency rate. Still, a considerable part of the variation in the QOL dimensions cannot be explained simply with price, ownership type and dependency rate.