Background Physicians should be engaged in quality-improvement activities to make the systems in which they work safer and more reliable. However, many physicians are still unable to contribute to patient safety initiatives that lead to safer, high-quality care for their patients.
Objective To survey 10 high-performing hospitals in the USA to determine how they engage their physicians in quality and safety.
Design Qualitative study that used site visits and a semistructured 20-question interview.
Setting Ten high-performing US hospitals were chosen from the 2010 US News and World Report Best Hospitals and the Leapfrog Group on Patient Safety.
Participants Forty two interviews were conducted with forty-six quality leaders including CEO's, Chief Medical Officers, Vice Presidents for Quality and Safety and physicians.
Measurements Site visits and in-person interviews were conducted during 2010–2011. The interviews were transcribed and coded using the constant comparative method for further analysis by the team.
Results The authors developed a six-point framework for physician engagement in quality and safety as a constellation of the best strategies being used across the country. The framework consists of engaged leadership, a physician compact, appropriate compensation, realignment of financial incentives, data plus enablers and promotion.
Limitation The qualitative design and the small number of hospitals surveyed mean that the results may not be generalisable.
Conclusion There remain many ongoing barriers to physician engagement in quality and safety. Some high-performing hospitals in the USA have made significant improvements in engaging their physicians in quality and safety. The proposed framework can assist organisations in the development of strategies to engage physicians in quality-and-safety activities.
- graduate medical education
- health policy
- healthcare quality improvement
- information technology
- evidence-based medicine
- patient safety
- diagnostic errors
- continuous quality improvement
- financial incentives
- lean management
- qualitative research
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
- graduate medical education
- health policy
- healthcare quality improvement
- information technology
- evidence-based medicine
- patient safety
- diagnostic errors
- continuous quality improvement
- financial incentives
- lean management
- qualitative research
Physicians should be engaged in quality-improvement activities to make the systems in which they work safer and more reliable.1–3 Prior evidence demonstrates that safety and quality improve when physicians are engaged and committed to the system in which they work.4–6 However, most physicians are ill equipped to lead patient safety initiatives, and many physicians struggle to optimally contribute to patient safety and quality-improvement efforts that lead to safer, high-quality care for their patients. This often results in quality-and-safety activities being delegated to others working in the health system.7
One of the impediments to greater success in quality improvement may be the underutilisation of physician skills.8 Physicians engaged in quality-and-safety activities bring a unique perspective and skill set, being able to focus on patient outcomes and inspire colleagues to improve care for their patients.9 While many organisations concede that physician engagement in quality and safety is important, it remains difficult to define, measure and improve. It is therefore not surprising that so few health systems have articulated a sustainable plan for engaging physicians in quality and safety.5 This barrier represents a significant challenge for health-system leaders seeking to achieve measurable improvements in patient care.4–6 ,10 ,11
While physician engagement is not widespread across healthcare systems, there are well-documented high-performing organisations that do engage physicians in quality and safety successfully.5 The goal of our qualitative study was to assess the themes common to organisations with high physician involvement and to explore how these high-performing systems engage their physicians in quality-and-safety activities with a focus on identifying the key facilitators and barriers to physician engagement.
Study design and sample
We selected a convenience sample of high-performing US healthcare organisations from among the 2010 US News and World Report survey on Best US Hospitals12 and The Leapfrog Group on Patient Safety (table 1). While these surveys have methodological flaws, they represent a useful starting-point to identify hospitals with potentially high levels of physician engagement in quality and safety. From these lists, 11 institutions representing adult and paediatric care were approached on the basis of recommendations from health-policy experts who considered the institutions to be leaders in the area of physician engagement. Some, but not all, hospitals were ranked in the honour roll of the US News and World Report, and a smaller number were among the top 65 hospitals identified by the Leapfrog Group on Patient Safety. Only one institution contacted declined to participate. At each of the 10 participating institutions, we asked the Chief Executive Officer to identify key leaders in quality-and-safety activities within the organisation that might participate in interviews. A total of 46 leaders in quality-and-safety improvement were interviewed during 42 interviews, with some of the interviews including more than one person. Coding and data analysis was based on 42 interviews. The interviews ranged from three to six interviews per health system, and included a mix of Chief Executive Officers, Vice Presidents for Quality and Safety, Chief Medical Officers, and practising physicians (table 2). The study was reviewed and approved by the Partners Human Research Committee.
A semistructured 20-question interview was developed by study team members, focusing on collection of interview demographics, experience and training in quality and safety, positive and negative examples of physician engagement, and barriers and costs of physician engagement. We also queried respondents regarding physician training programmes, performance reporting and measurement of physician engagement in quality and safety (web-only supplementary file).
The interview instrument was developed and piloted at Partners Community Healthcare (PCHI), the physician network for Partners HealthCare System in eastern Massachusetts. Following this initial set of interviews, subsequent interviews were completed with all participating institutions during a 4-month period from 2010 to 2011. Two of the institutions were hospitals within Partners HealthCare System, but their physician management infrastructures were distinct from those of PCHI. All interviews were conducted face to face by JT, digitally recorded and transcribed using encrypted software. All interviews were confidential and were typically 30–45 min in duration.
The transcripts were analysed using the constant comparative method. This method allows concepts from coded data to be extracted and compared with subsequent interviews to develop recurring themes.13 ,14 The transcripts were manually coded by a single author (JT) following a review of a subset of the transcripts by the team and agreement on the coding scheme by the other authors. Most of the emergent themes that arose from the interviews fell within the engagement framework. Themes around quality-and-safety educational programmes, training of quality-and-safety leaders, primary versus specialist physician engagement and health reform did not fall directly within the framework. These themes were coded and analysed but excluded from the discussion. Once all the transcripts were coded, the authors reviewed the themes to ensure agreement.
Role of physician engagement
The definition of physician engagement in quality and safety was generally consistent across all sample hospitals. Most respondents defined engagement as physicians working to reduce unjustifiable variation in care, considering the processes and systems in which they care for their patients. One Chief Executive Officer reported that physician engagement is ‘about physicians owning the most optimal way in which healthcare is delivered so that it is focused, smooth, effective, and achieves desired patient outcomes.’ Other responses stressed that physician engagement means purposeful, constant commitment to delivery of organisational objectives around quality and safety. Another consistent theme in the interviews was that, among healthcare providers, physicians have the most influence on variation in healthcare outcomes—and without engagement and alignment of physicians, there is no meaningful way to influence variation in healthcare delivery. No organisations identified use of a reliable tool to measure physician engagement.
Key drivers of physician engagement
The key drivers of physician engagement in quality-and-safety activities were consistently described by six themes (table 3). Not all health systems were successfully implementing all aspects of the framework; some systems were stronger in some areas than others (table 4).
Leadership was the most frequently cited component within the overall framework as a key driver of physician engagement, with 70% of interviewees endorsing its importance. Interviewees focused on the alignment between physicians and the organisation on quality as a shared goal. One Chief Medical Officer summarised the essence of leadership: ‘It is vital to have leaders with street credibility that are able to get physicians to buy into quality and safety and develop a culture that communicates a clear institutional mandate for quality and safety around patient outcomes.’
The physician compact represents the rules of engagement between physicians and the institution, outlining mutual expectations.15–17 A physician compact was mentioned in 15% of interviews as a key driver of physician engagement. Four of the sample health systems were in the process of developing a compact (n=3) or already had an explicit compact in place (n=1). For this latter organisation, the chief executive officer described the process: ‘The compact took 12 months of deliberation before both sides were happy with its contents, and it is still reviewed at monthly meetings to ensure both parties’ behaviour is consistent with the compact.'
A compact was described as essential to bridging the gap between the old concept of autonomous physicians protected by the organisation to the new aim of high quality, safe care focused on team work and patient outcomes. One chief medical officer emphasised this concept: ‘Physicians need to realise that tangible and intangible rewards are aligned with achieving the compact goals and that the focus is the patient and not the physician.’
Adequate compensation for time dedicated to quality-and-safety activities was mentioned in 50% of interviews. Interviewees reinforced the notion that physician schedules are completely booked for routine patient care, limiting time for engagement in quality-and-safety activities and leading to frustration by both physicians and health system leadership. One interviewee commented: ‘Too many hospital leaders believe that physicians can do quality at tea time and between surgical cases and that quality activities are simply an additional component to be added onto a physician's schedule.’
Financial compensation was cited as a means to protect time for physicians to adequately participate in quality-and-safety activities, either as part of their regular salary or through extra payment in a fee-for-service environment. This process was viewed as one method to hold physicians accountable for working to achieve improved quality and patient outcomes.
Realignment of financial incentives
There was significant divergence in the views of interviewees regarding the appropriate role of performance-based financial incentives as a means of engaging physicians in quality-and-safety activities. Approximately one-third (33%) supported the use of such incentives, while 31% directly opposed their use. This was in contrast to the general support outlined earlier for compensating physicians for their time spent outside direct patient care participating in quality-and-safety activities.
Some leaders strongly supported the use of modest financial incentives to change physician behaviour, while others questioned the role of such incentives. As described by one chief quality officer: ‘… Financial incentives don't change a physician's behaviour and are indeed an insult to the professionalism of physicians.’
This strong negative view was not universally held, as five of the sample health systems had financial incentive programmes in place. In the words of one interviewee: ‘…incentives are not about paying physicians to work longer and harder but rather about a positive incentive for doing the right thing in accordance with institutional priorities.’
Nearly two-thirds (64%) of respondents stated that if financial incentives were to be used, they should be based on outcome measures rather than process measures.
Data and other enablers
Data on clinical performance and safety were cited by 66% of interviewees as an important driver of physician engagement. Health system leaders felt that these data were very helpful in convincing physicians of the existence of gaps in quality and safety, as well as an opportunity to encourage healthy competition among physicians to improve their patient care outcomes. According to one Vice President for Safety and Quality: ‘Physicians are driven by data and will respond to peer-pressure in a structured format, especially if they are seen to be under-performing in relation to their peers.’
One important theme was the value of active engagement of physicians in the process of developing performance metrics, focusing on whether these metrics were acceptable, defendable, clinically meaningful to patients and readily measured with existing resources. Involving physicians in this process was viewed as a method of supporting their professionalism, a concept endorsed by 20% of interviewees as an important component of engagement. One chief executive officer remarked: ‘By over-emphasising financial incentives health leaders have neglected the inbuilt professionalism of physicians whose search for personal excellence is a core part of their professional values.’
Interviewees also commented on whether data should be published internally or be made public and published externally. Nearly all (95%) interviewees agreed on complete internal transparency with identification of individual physicians. However, only 52% were supportive of publishing these data externally, and 26% were solidly against such external publication. Consistent with these divided views, only five of the sample health systems were publishing de-identified data externally. One chief medical officer commented: ‘It is the internal data that drives culture change by identifying the best and the worst performers and learning from the best.’
In addition to reporting on clinical performance, it was consistently noted that data need to be accompanied by processes or tools that enable physicians to take action to improve patient care. Potential enablers identified by interviewees included development of clinical practice guidelines, electronic reminders and patient registries. One quality officer remarked: ‘Enablers are important as physicians don't just need appropriate data; they also require enablers or processes to improve their performance.’
Traditional academic promotion is based on a physician's research, clinical, and teaching productivity, with no systematic consideration of physician involvement and participation in quality-and-safety activities. Some interviewees (14%) reported taking substantial steps to incorporate physician engagement in quality-and-safety activities as a legitimate criterion for academic promotion, with the goal of encouraging and rewarding these activities. As a Vice President for Safety and Quality noted:
Quality grants should be peer reviewed within the institution and quality-and-safety publications should be considered on a par with traditional research publications. This may also lead to academically minded staff being attracted to institutions that support the pursuit of quality and safety in a scholarly manner.
Barriers to physician engagement
Many barriers to physician engagement in quality-and-safety activities were identified (table 5). Lack of physician time was the most frequently cited barrier, reported by nearly one-half of interviewees. Other key barriers were institutional culture (31%), physician desire for autonomy (17%), insufficient training in medical school (17%) and general lack of quality-improvement skills (13%). As summarised by one chief quality officer: ‘The tradition of physician centric practice that values independence over teamwork leads to physicians being resistant to change and lacking the necessary skills to contribute to quality and safety activities.’
The cost of physician engagement is also an important potential barrier. All interviewees recognised that this cost is part of the institutional commitment to achieving quality at the highest level of their organisation. One quality officer pointed out that: ‘Most importantly the cost of not engaging physicians in quality and safety is much higher than the cost of engaging them.’
However, most interviewees were unaware of the exact financial cost of physician engagement in quality and safety, through either direct programme costs or the indirect costs of removing physicians from their clinical practice to participate in quality-and-safety activities. Anecdotally, some hospitals estimated their return on physician engagement to be in the region of 4:1 to 6:1 for every dollar spent. One hospital noted a 50% reduction in its medical malpractice insurance premiums since the inception of a quality-and-safety programme.
We conducted a qualitative evaluation of physician engagement in quality-and-safety activities at 10 leading health systems across the USA and identified six key drivers of successful engagement. We would note that our proposed framework does not require the concurrent presence of all six drivers to engage physicians in quality-and-safety activities. Our analyses showed that all six drivers were not in place at any one institution, highlighting both the potential challenge to successfully implementing a broad-based approach to physician engagement and the potential to achieve engagement with narrower, yet appropriately targeted strategies.
While our proposed framework is not intended to represent a strict sequential implementation, two key components deserve highlighting as important factors early on in the physician-engagement process. First, while no health systems were employing all elements of the proposed framework, all reported highly engaged leadership committed to improving quality and patient safety. This leadership support is likely prerequisite to any subsequent activities to engage physicians successfully. Second, the development and use of an explicit physician compact can be used to facilitate subsequent discussion around compensation, incentives, data handling and academic promotion. Only four of the health systems in our study were explicitly using a physician compact, highlighting a key area where even high-performing systems might achieve substantial gains in physician engagement.
We found a great degree of divergence regarding the use of financial incentives to engage physicians. The literature on financial incentives altering physician behaviour remains mixed.18–21 Most current financial incentives focus on process measures that may not resonate with physicians' views of quality and safety, leading to a disconnect between the practitioners and leaders driving the quality-and-safety agenda. Health-system leaders and physicians may receive financial incentives more positively and respond more consistently if a collaborative process is undertaken between these parties to ensure alignment around core professional values.
We noted a move towards increased transparency of performance reporting among some of the sample health systems, with five publishing deidentified data externally. However, the embrace of such transparency was not universal, and many interviewees were cautious in their support for publishing data externally due to concerns about the robustness of the data and the potential damage that misleading data could cause to the health system and its physicians. Performance reporting and transparency are often regarded as critical components of transforming healthcare. However, it is not clear whether public transparency is more effective than internal publishing of performance data with regard to engaging physicians in quality-and-safety activities.22 ,23 This is an important area for future exploration, particularly as these data indicate some contention over this issue.
We included the importance of linking academic promotion to quality-and-safety activities as part of our framework, recognising that this concept may not apply to all physician groups across the country. However, our interviews indicate that this driver resonates strongly with physicians at large academic medical centres. The development of an effective synergy between promotion and engagement in quality-and-safety activities represents a lever that can drive the generation of new knowledge in this field, as well as increase the legitimacy and support of academic work in quality-and-safety improvement.7
Despite physician engagement in quality and safety being seen as a key driver of high-performing organisations, none of the surveyed health systems are measuring physician engagement in quality and safety. Almost all interviewees felt it would be helpful to standardise measurement of physician engagement in a manner that facilitates benchmarking with other organisations and across physicians within their organisation. To our knowledge, no such reliable tool exists. Some health systems were using indirect measures of physician engagement such as patient experience of care surveys or culture of safety surveys. We feel that more research is needed to develop a more direct measure of physician engagement in quality and safety.
In the majority of the hospitals that we surveyed, the physicians are salaried employees of the hospital. This likely facilitates, but does not guarantee, physician engagement in quality and safety. It allows hospitals to allocate a portion of a physician's time specifically dedicated to quality-and-safety activities. It also allows for more flexibility in aligning financial incentives and direct use of hospital performance data in measuring individual physician performance. These activities may be more challenging to implement when physicians are self-employed or employed independently of the hospital.
We believe that our framework for engagement offers a potentially actionable plan to improve physician engagement in quality and safety. It will be important to apply our framework to lower-functioning hospitals, and to determine if implementing key pieces of the framework are associated with improved engagement and ultimately improved quality and patient safety.
Our study is strengthened by the inclusion of a multiple high-performing health system using a range of physician employment models and representing many different geographic healthcare markets. However, our findings should be interpreted in the context of our study design. First, our engagement framework may not generalise to community hospitals. Academic promotion and financial compensation for quality-and-safety activities may not appeal to these community hospitals and their physicians. However, we feel that other aspects of the engagement framework such as engaged leadership, correctly aligned financial incentives and performance data can drive physician engagement in quality and safety at smaller, non-academic hospitals. Second, our convenience sampling strategy did not include hospitals with potentially less physician engagement, and we cannot confirm that the key features of engagement identified in our study are unique to high-performing institutions. Third, while we included the views of practising physicians, our sample was weighted towards senior leadership, and did not include other key members of the quality improvement process such as nursing. Future research should also include a focus on the views of these important members of the healthcare-delivery team.
There are many ongoing barriers to physician engagement in quality and safety. Some high-performing US healthcare systems have made significant improvements in recognising that physicians must be engaged to support transformation in healthcare that improves quality and patient safety. We have presented a framework for physician engagement representing a constellation of the best practices we encountered across the country. Healthcare systems seeking to engage their physicians will benefit from implementing as many of the key drivers as possible, while also considering the significant barriers identified.
Additional materials are published online only. To view these files please visit the journal online (http://dx.doi.org/10.1136/bmjqs-2011-000167).
Funding This study was supported by the Commonwealth Fund Harkness Fellowship in Health Care Policy and Practice.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.