What do hospital decision-makers in Ontario, Canada, have to say about the fairness of priority setting in their institutions?

Priority setting, also known as rationing or resource allocation, occurs at all levels of every health care system, including governments, funded provincial/territorial agencies, pharmaceutical benefit-management organizations, hospitals and clinical programs [1]. Countries with very different health care systems and levels of health care are all grappling with the problem of how to reconcile growing demands and constrained resources [2]. Hospitals, in particular, are struggling to meet growing demands, affordably, without compromising delivery of services [3, 4].

Daniels and Sabin have proposed a framework for priority setting in health care institutions called 'accountability for reasonableness' [5‐7], which links priority setting to theories of democratic deliberation, operationalizing the ethical concept of fairness. Fairness is a key goal of priority setting. According to 'accountability for reasonableness', health care institutions engaged in priority setting have a claim to fairness if they satisfy four conditions of relevance, publicity, appeals/revision, and enforcement (described in Table 1).

'Accountability for reasonableness' may be an effective framework for describing the components of priority setting, evaluating the fairness of priority setting in hospitals and identifying 'good' practices and opportunities for improvement. It can help hospitals improve their priority setting practices and can be an effective driver of health care reforms [8‐10].

'Accountability for reasonableness' has been used to evaluate priority setting in health systems [11]. However, only a limited number of studies have addressed how priority setting occurs in hospitals [12‐17], and no study has attempted to survey the views of hospital decision makers throughout an entire health system about the fairness of priority setting in their institutions.

The purpose of this study is to elicit hospital decision-makers' self-report of the fairness of priority setting in their hospitals using an explicit conceptual framework, 'accountability for reasonableness'.

Overall, 60.7% of respondents indicated their hospitals' priority setting was fair, while 79% stated their hospitals met their priority setting goals. With respect to the 'accountability for reasonableness' conditions, respondents indicated their hospitals performed best for the relevance (75.0%) condition, followed by appeals/revision (56.6%), publicity (56.0%), and enforcement (39.5%). (Refer to Table 3).

Respondents rated the relevance ( = 3.94) condition significantly higher than the publicity ( = 3.57), appeals ( = 3.52), and enforcement ( = 3.25) conditions (p < .05) (paired-samples t test). While publicity and appeals conditions were not significantly different, respondents rated these conditions significantly higher than the enforcement condition (p < .05). The distribution of ratings suggested that there was ample opportunity for improvement, with most room for improvement in meeting the enforcement condition.

Priority setting and fairness ratings were positively correlated (r = .51, p < .01). Fairness ratings were also positively correlated with each of the 4 'accountability for reasonableness' conditions (p < .01), as were meeting priority setting goals (p < .05).

The bivariate correlation between each of the accountability variables of relevance, publicity, appeals/revision and enforcement was strongly positive (p < .01). The Pearson correlation coefficients among the 4 conditions ranged from 0.41 to 0.57. The internal consistency (Cronbach's α) of all of the 'accountability for reasonableness' conditions was very good (α = .78).

Bed size (r = -.24, p <.05) and operating budget (r = -.32, p < .01) were negatively correlated with the rating of fairness in priority setting.

To our knowledge, this is the first published survey of hospital decision makers covering an entire health system, using Daniels and Sabin's framework of 'accountability for reasonableness'. It provides data to understand how fair priority setting processes are, what decision makers' priority setting goals are, and how well these goals are met.

Several studies, using either survey or qualitative methodologies sometimes in combination with hypothetical "tradeoffs", have examined priority setting from theoretical perspectives aimed at capturing the public or professional's preferences or values in priority setting [21‐24]. Relatively few studies have described what is happening in real-world contexts with a view of evaluating and improving priority setting activities [25‐31]. These have described actual priority setting in individual hospitals, focusing on strategic planning [12], surgery [13], critical care [14], new technologies [15], and the rationing of new drugs [17], with no study reviewing priority setting in hospitals across the health system.

This study makes five contributions to knowledge on priority setting. First, there is ample room for improvement in fair practices within Ontario hospitals as described by their Chief Executive Officers. Although many hospital CEOs felt that their priority setting was fair, ample room for improvement was noted, especially for the enforcement condition, followed by publicity, appeals/revision and relevance. While 79% of decision makers felt their hospitals met their priority setting goals, only 60.7% felt their processes were fair. According to decision makers, the perceived 'gap' was greater in meeting fairness than in priority setting goals.

Second, it is feasible for decision makers to assess the fairness of priority setting in their institutions on a quantitative basis according to the 'accountability for reasonableness' framework, with greater than half of respondents saying their processes are fair. There is a high degree of association between 'fairness' and the internal components of 'accountability for reasonableness', lending support that fairness can be operationalized through 'accountability for reasonableness'. As well, the internal conditions of 'accountability for reasonableness' are highly related, and positively correlated, suggesting relevance, publicity, appeals/revision, and enforcement are measuring comparable aspects of fairness. Consistent with this, the high Cronbach's α finding is promising for the development of an 'accountability for reasonableness' scale of perceived fairness of priority setting in health care institutions. Measuring decision makers' self-report of their perception in meeting the various conditions of 'accountability for reasonableness', the scale would provide a practical tool for decision makers to assess self-reported views over time, noting progress in meeting conditions, while providing a reference point for needed improvement (i.e. enhance meeting 'publicity' condition). Finally, the finding of a negative relationship between number of hospital beds or budgets and self-report on fairness is consistent with the view that smaller hospitals may be perceived to have fairer processes, with greater involvement of their local communities in decision making and emphasis on transparency and "trust".

Third, decision makers' views were surveyed at a health system level to determine what their actual priority setting practices were. Four complex, interrelated priority setting goals were described, suggesting a balance of need, quality, budget and organizational goals in priority determination. In addition, the study points to a close alignment of factors required in the meeting of priority setting goals and in the elements of fairness. In providing decision maker perspectives on characteristics of fairness, this study also adds to previous empirical work suggesting fair priority setting depends on a fair priority setting process [31]. Additional research is required to understand the capacity of organizations to improve such practices.

Four, this study has shown, from a blend of qualitative and quantitative approaches, that it is feasible to operationalize the 'accountability for reasonableness' framework, through the design of survey instruments to facilitate self-evaluation, and identification of good practices. This data can be shared with decisions makers to improve the fairness of their priority setting processes in meeting the four conditions of 'accountability for reasonableness'.

Finally, this study expands on the likely relationship between leadership and priority setting in which leadership was found to contribute to perceptions of fairness in two committees engaged in priority setting for new health care technologies. Study results indicate that greatest room for improvement exists in meeting the 'enforcement' condition, with decision-makers describing critical leadership success factors, including: the role of governance in establishing policy and in meeting this condition, the need to set achievable corporate goals, and the significance of a lobbying funding function. Further study is needed to clarify the nature of this leadership contribution.

In this first survey of Chief Executive Officers within a health region reporting on their assessment of the fairness of priority setting practices in their institutions according to 'accountability for reasonableness', ample room for improvement was observed, especially for the enforcement condition. Additional study is required to understand how application of 'accountability for reasonableness' will vary within and across institutions, and is shaped or influenced by various parameters, including the nature of corporate leadership. These evaluations can help to identify good practice opportunities for improvement that can be shared between local institutions or used within government to drive health care reforms.