Introduction

The use of clinical quality registries (CQRs) is frequently emphasised as a means for continuous quality improvement.1 2 By collating standardised information on clinical care processes and patient outcomes within demarcated areas of healthcare and making feedback data available to the participating sites, the use of CQRs can provide the basis for improving suboptimal practice.3

However, to achieve the purpose of quality improvement, the use of CQRs must be thoroughly implemented, including active use of the collected data for follow-up and learning opportunities. Studies indicate that this may often not be the case. A recent systematic review found that only a few high-quality studies have been able to show an effect in terms of improved quality of care.4 Despite substantial investments into increased use of CQR data in Sweden, national evaluations have shown that the registries have not been drivers of local quality improvement, with data use often being limited.5 On the other hand, the application of data in local quality improvement work may differ between registries,5 6 indicating data use may be registry and context-dependent.

Some of the determinants for use of data include: data relevance,7 8 perceived quality of data, timeliness of feedback,9 know-how among staff,10 sufficient resources,5 10 collaboration between relevant organisational tiers6 7 11 and engagement of both frontline staff and managers.12 Thus, while collaboration and engagement across the organisation is important, it is unknown whether staff and managers are influenced by the same determinants, and whether they share perceptions on the use of CQR data. Studies of other types of quality improvement initiatives have suggested that determinants for use of data may differ between these two occupational groups, and that managers may have more overall positive views of the initiatives compared with frontline staff.13–15 As such disparities have possible implications for the use of data,12 13 studies in the field of CQRs seem warranted.

CQRs are typically introduced in clinical areas where there is a gap between evidence and practice.3 One such area is cardiac rehabilitation, which despite being an important part of secondary prevention for patients recovering from heart disease,16 is consistently reported to have evidence-practice gaps.17 18 CQRs with the purpose of monitoring and improving cardiac rehabilitation services have been established in at least seven countries.19 Although sizeable resources are invested into development, administration and data collection of these CQRs,3 19 it remains unclear to what extent the data are being used and what drives the use of data for local quality improvement among registries in this clinical area. The need for knowledge on these aspects is underscored by the fact that data-driven quality improvement currently is high on the healthcare agenda, with CQRs potentially having significant roles in delivering data.2 20 21

The objective of this study was to investigate the use of data from a CQR for cardiac rehabilitation in Denmark, considering the extent to which data are used for local quality improvement and what facilitates the use of these data, with a particular focus on whether there are differences between frontline staff and managers. The CQR in question represents a nationwide registry based on international evidence, intended to be used primarily for quality improvement.22

Methods

Results

Participants

Out of the 30 departments, 28 were represented in this study. A study flow diagram is presented in figure 1. The survey was sent to 175 individuals of whom 101 responded (58%), of which 62 were frontline staff, 19 mid-level managers and 20 heads of departments (table 1). Response proportions differed among the different groups of staff, with 78% of frontline staff responding and 35% of heads of departments. Characteristics of respondents and non-respondents are presented in table 1.

View this table:

• View inline
• View popup

Table 1

Characteristics of respondents and non-respondents

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Study flow diagram.

Most respondents were female (87%), aged 41 years or older (84%), with three or more years of experience with their role in the registry (68%). Among non-respondents, 33 (mainly managers) reported not having sufficient knowledge concerning the registry to respond to the survey questionnaire. No other reasons for non-response were reported, although two managers stated that they, besides insufficient knowledge, did not have enough time.

Discussion

This study sought to survey the use of data from the Danish Cardiac Rehabilitation Database to determine the extent to which data are used for local quality improvement and what facilitates the use of these data. Findings indicate that data from the registry were used in local quality improvement work to a relatively limited extent. Frontline staff on average reported using data for one of seven suggested purposes, while heads of departments reported using it for three. It was not possible to distinguish between high and low extent of data use per se, as there is no shared understanding of standards for use of CQR data yet. It may be that such standards are unrealistic to establish and that use of data should always be judged based on the individual case. In the current study, we regard the use of data to be relatively low as only 29% of frontline staff and 50% of heads of department in the present study reported to carry out data-driven improvements based on results in the registry, and less than one in three present results to members of staff. Furthermore, in a previous study we identified gaps between evidence and practice in nearly all departments providing cardiac rehabilitation in Denmark.18

While the literature regarding use of CQR data is sparse, our findings are comparable to the findings of Fredriksson et al.6 They studied the use of data in local quality improvement, according to physicians and managers in three Swedish CQRs (stroke, gallstone surgery and lung cancer), using the original version of the QWAQ. They found similar levels of reported use of data in the gallstone surgery and lung cancer registries, while it was higher in the stroke registry. The latter is considered to be more developed in terms of feedback with national benchmarks and validation of data.6 Such a degree of maturity has been suggested as a possible explanation for differences in use of data between registries.5 6 27

In the present study, a specific focus was on whether frontline staff and managers’ perceptions of data use differed. Frontline staff reported use of data on a department level to be lower than that of their managers. We found no studies investigating the perceived extent of use of CQR data among frontline staff as a separate group, nor was it possible to establish which (if any) of the two groups were ‘right’, that is, whether responses correspond to objective measures of data use. In a previous study of a patient safety programme, managers were found to hold a more positive view of the effectiveness of the initiative than frontline staff.13 Similar to the proposed explanation in that study, the different perceptions between users of the Danish Cardiac Rehabilitation Registry could be explained by managers’ greater overview, having insight into quality improvement initiatives across the organisation.13 However, it could also be that managers simply do not know as much about registry use in practice as their frontline employees, resulting in overly optimistic estimates. Yet another possibility is that the managers’ responses may be more influenced by social desirability bias, that is, over-reporting of desirable behaviours. Data-driven quality improvement is high on the healthcare policy agenda and managers (particularly heads of departments) are likely aware that their organisation is being benchmarked against others in annual reports and in other national, publicly available data such as the National Healthcare Quality Programme.1 Frontline staff, on the other hand, may not be faced with such pressure to apply data. Instead, they often focus on entering data.7 The differing perceptions between frontline staff and managers underscore the importance of accounting for perceptions of both groups when conducting studies of quality improvement initiatives13 15 to obtain a nuanced view from stakeholders in different positions.

While reported data use was positively associated with the quality and usefulness of data, management requesting data and personal motivation, none of these aspects were rated highly by the respondents. These aspects have previously been documented as influencing effectiveness of feedback from registries9 and thus would seem crucial to address and improve in order to maximise the chances of usage of registry data. The negative perceptions of data quality underscores the challenges of creating a CQR that captures the essential aspects of care in relevant quality indicators.3 9 This may be particularly difficult in a field such as cardiac rehabilitation, where a large proportion of the clinical intervention centres on lifestyle changes and improving quality of life, and where socioeconomic vulnerability among patients may add specific barriers to the perceived validity of data.28

In agreement with previous Swedish findings,12 resources, such as time and competencies, and receiving support were not associated with use of registry data. This may seem surprising, as it has been highlighted as potentially important by implementation frameworks such as the Consolidated Framework for Implementation Research26 and contradicts previous reports.5 10 29 Resources and support were also associated with use of data in univariate analyses (data not shown). Hence, it appears that resources and support are reduced in priority when compared with other aspects, reflecting a complex interplay of factors that influence the extent of data use. This complexity increases further when considering that different aspects seemed to be important for data use among frontline staff and managers.

The QWAQ covers a range of facilitators for use of data,12 however, all potentially relevant facilitators cannot be investigated. This is a common limitation of survey research, where questionnaire developers must weigh precision against the response burden. Still, our survey has yielded important insights into the relative importance of different facilitators.30 Further explanations concerning the use of CQR data can be derived from our previous qualitative studies in the cardiac rehabilitation field. For example, we have found that feedback data may not reach the frontline staff because it fails to pass through complex delivery pathways, staff may not know that local feedback data exists and a culture supporting quality improvement may not have been established31 (Helmark et al; article accepted for publication). Other plausible explanations, suggested by our previous work and supported by other scholars, are that roles and responsibilities for acting on data are unclear and that there is a general lack of time and understanding regarding the use of CQRs in improvement work in healthcare.15 27 31 Furthermore, clinicians tend to have their own perceptions of what constitutes quality of cardiac rehabilitation and may dismiss the defined indicators.31 32

We regard it as an important finding that nearly half of the survey non-responders stated that they could not respond due to lack of knowledge of the registry. The majority of these reports came from managers, particularly heads of departments. Although it cannot be expected that managers have detailed knowledge of CQR use in daily practice, policy documents emphasise that they should have a strong focus on data-driven quality improvement and intervene when quality targets are not met.25 Knowledge of the registry including how to apply data in improvement work is important to be able to link efforts to mission and strategic objectives, to allocate responsibilities and secure resources and to motivate employees.33 It has been highlighted previously that managers often have inadequate knowledge and spend too little time on quality improvement to meet these important objectives.33 34

Conclusion

This survey study among frontline staff and managers employed in clinical departments participating in the Danish Cardiac Rehabilitation Database indicate a relatively limited use of data from the database, where frontline staff reported use to be lower than that of their managers. Factors associated with use of data were the perceived data quality and usefulness, management request for data and personal motivation to use data. A difference between managers and frontline staff was found, as data quality and usefulness was important for managers' reports of data use, while frontline staff reported use to be associated with their own motivation and with management involvement in quality improvement work. These findings suggest that a complex interplay of factors is associated with use of CQR data, with different aspects being important to different types of users. Furthermore, it emphasises the need to include both managers and frontline staff when evaluating use of CQRs.

Although translation of the results from this study to other registries and settings should be done with caution, our results combined with the body of literature in the implementation science field suggest that while in the planning process of a new registry, the quality of the CQR and the readiness to receive it in practice should be carefully evaluated. Future studies should evaluate initiatives to enhance cardiac rehabilitation registries' data quality and relevance, and to build quality improvement capacity among clinical teams and their managers with regard to applying CQRs.