The main contributions of this study to the literature on HRM and performance in the health care sector concerns the use of a multidimensional performance perspective. In this respect, we examined three different outcomes: financial (net margin), organizational (client satisfaction), and HR (sickness absence). The analysis includes job satisfaction, which can be regarded as a ‘black box’ variable: a mediating variable connecting HR practices and performance.
The results confirm the basic notion that HRM and performance within the health care sector are linked. Our final SEM model shows that HRM is - directly or indirectly - linked to all three outcomes. When organizations apply - according to their employees - more HR practices, this is associated with greater client satisfaction, less sickness absence, and a better net margin. With respect to organizational and HR outcomes, the hypotheses regarding the mediating effect of job satisfaction are confirmed. This is in accordance with the perspective that employee attitudes are an important component of the ‘black box’ between HRM and performance. In this respect, our study showed that higher job satisfaction is associated with higher organizational performance. More specifically, in line with the assumption, our research showed a positive association between employee satisfaction and customer satisfaction because if employees are satisfied with their jobs, they are likely to behave toward customers in ways that yield positive service experiences. A more extensive use of HR practices leads to more satisfied employees. This greater satisfaction ‘reflects’ on the clients, as satisfied employees will do more for them . Moreover, satisfied workers are less likely to call in sick than less satisfied workers.
HR practices are directly related to financial outcomes, although the explained variance is small. Furthermore, we found that job satisfaction does not mediate the relationship between HRM and net margin. As we mentioned in the introduction, financial outcomes are a distant outcome of HRM. In fact, the literature about strategic management informs us that organizations can use different strategies to achieve their objectives . In addition to a high performance strategy, organizations can also employ a low cost strategy . Boxall and Purcell  describe the ‘mass service market’ - which includes care - as a ‘service market with some quality differentiation’. Organizations can follow various strategies to become (financially) successful. One possible strategy implies investing in employees, which will likely result in more satisfied employees. Another strategy implies cutting costs, which will result in reduced investments in employees and (most likely) less satisfied employees. The finding that HRM has a direct effect on financial outcomes may be because a low cost strategy also implies the use of certain HR practices, for instance performance management. It can thus lead to financial success without positively affecting the satisfaction of employees.
We conclude this article by presenting some limitations. An important limitation of this research - but also of many other studies in this area - is the hidden assumption that the same mix of HR practices will work for all organizations. Therefore, the inclusion of HR strategy in research designs will be an important addendum.
The possibility of considering various data sources (employee, client and ‘objective’ performance data) is an important - and unique - advantage of this study. However, it also has some drawbacks. The scales used are not based on previous academic literature. In further research, validated scales should therefore be employed. Moreover, a disadvantage of using secondary data is that not all the desired research concepts were covered in the data.
A further limitation is the sample size. Although the underlying dataset is large, the data were aggregated at the level of 85 health care organizations. This could be considered quite low. However, Bentler and Chou  recommended a ratio of sample size to free parameters of at least 5:1. In our analysis, the model tested was simple, and the ratio of the number of free parameters to the number of cases did not fall below under 5:1. Related to this, several studies using SEM with a small sample size are available [47–49]. Nevertheless, future studies might attempt to replicate the findings using larger sample sizes.
Furthermore, the results of this study should be interpreted in light of the study’s context and sample. The study was conducted in the Netherlands, which features a social health insurance scheme in health care financing and a mix of public and private provider organizations in health care provision . This is in line with other ‘Bismarck’ countries, such as Belgium, Germany and France . It would be interesting to replicate our study to test the proposed model in other countries using different kinds of health care systems.
In conclusion, our empirical results underscore the importance of HRM in the health care sector. We can state that HRM makes a difference, especially for HR and organizational outcomes. Its impact on financial performance is less strong. Job satisfaction links HR practices and organizational and employee outcomes. In conclusion, further analyzing HRM in the health care sector will be a productive endeavour for both researchers and practitioners.