Article Text

Download PDFPDF

High cost of broken relationships
  1. Christine A Sinsky1,
  2. James Morrow2
  1. 1 Professional Satisfaction and Practice Sustainability, American Medical Association, Chicago, Illinois, USA
  2. 2 Granta Medical Group Shelford Medical Practice, Cambridge, UK
  1. Correspondence to Dr Christine A Sinsky, American Medical Association, Chicago, IL 60611, USA; christine.sinsky{at}

Statistics from

Request Permissions

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.

Turnover is costly. When primary care physicians (PCPs) leave their practice for another location, leave medicine altogether for another career, partially leave by reducing clinical working time or retire early then relationships are broken, and patients and payors pay a price. Costs are higher and quality is lower.

In the USA we found that patients who had lost their PCP incurred additional healthcare expenditures: $61 more per year for non-Medicare patients (typically those under 65 years old) and $189 more per year for Medicare patients.1 This results in an estimated $86 000 additional spending per departing PCP, totalling $1 billion per year across the USA in excess healthcare expenditures due to disruption in relationships between patients and their PCP. Patients receive more and unnecessary testing, with the consequential risk of iatrogenic harm, when evaluated by a physician who does not know them well or at a site of service with higher costs and a more test-heavy approach, such as a single specialty clinic, urgent care or the emergency department.

Comprehensive primary care has long been shown to be associated with better cost and quality outcomes in comparisons by country or by region.2 Continuity with one’s personal physician has also been associated with reduced health service utilisation, complications and mortality.3

The study by Parisi et al 4 in the current issue of BMJ Quality and Safety adds to the literature by further demonstrating the value of continuity and the costs incurred by patients and payors when continuity is broken. Using administrative data for general practitioners (GPs) in England from 2007 to 2019 researchers determined that patients cared for in practices with persistently high levels of physician turnover, defined as more than 10% GP turnover in each of three consecutive years, had significantly worse outcomes: less access to same-day care, less continuity with one’s personal physician, more emergency department attendances and more emergency (ie, unplanned) hospitalisations.

Perhaps even more concerning are the trends in continuity and access across all GP practices over the decade of the current study. The likelihood of seeing one’s preferred GP dropped from 72.5% to 56.1%, while the odds of achieving a same-day appointment plummeted from 80.3% to 32.3% over the same interval.4 Continuity and access to care appear to be at risk in the UK as they are in the USA.

The study by Parisi et al 4 did not address the impact of increasing numbers of physicians working less than full time, yet this is a substantial contributor to the effective reduction in the size of the physician workforce and is thus also a threat to patient access to care, quality and cost. Burnout, dissatisfaction with work–life integration and dissatisfaction with the electronic health record (EHR) are each associated with intent to reduce hours providing clinical care.5 6 In one longitudinal study, for every 1 point increase in burnout on a 7-point scale there was a 43% increase in the odds of reducing clinical hours.5

Given the importance of continuity and access in primary care, it is critical to understand why physicians leave a profession they have invested so much of their lives preparing for and which is inherently meaningful. Parisi et al found that practices with high patient complexity and reduced financial support were more likely to have persistently high turnover.4 This suggests an imbalance between job demands (patient complexity) and job resources (financial and clinical staff support) is a risk factor for turnover.

We now turn to burnout as a key to understanding and addressing physician turnover and its negative consequences for patients and payers. Longitudinal studies in the USA document that physicians who experienced symptoms of burnout are twice as likely to have left their organisation 2 years later compared with those who did not experience burnout at baseline.7

The mismatch between job demands and job resources is a significant driver of burnout, which in turn is associated with physicians’ intention to leave their current practice and intention to reduce clinical work. Job demands have increased for PCPs over the past 20 years. In 2003, it was estimated to require 18 hours/day for PCPs to accomplish all of the recommended prevention and chronic illness care.8 9 By 2020, and still before pandemic, these estimates had risen to 26.7 hours/day.10 There is simply more work to be done than can be done in the ‘physician-acting-alone’ model of care commonly found in primary care in the UK and the USA. It should not be surprising, then, that burnout and intent to leave are higher when there are low levels of staff support (ie, fewer job resources)11 and greater dissatisfaction with the tool used to organise care, the EHR.6 Poor values alignment, low work control, poor teamwork and a chaotic workplace are each also associated with intention to leave one’s practice.12

The COVID-19 pandemic has exacerbated job demands. For example, patient medical advice requests via the patient portal increased abruptly by 57% across the entire US Epic user base at the onset of the pandemic and have remained at this higher level since, despite a return to baseline levels of in-person visits.13 The pandemic also exacerbated burnout and intentions to reduce clinical hours or leave clinical practice. Sixty-three per cent of US physicians now experience symptoms of burnout, reversing a 6-year downward trend in burnout rates and reflecting a marked rise from 38% just 1 year earlier.14 Whereas in the first year of the pandemic three-quarters of US physicians would choose to be doctors again, by the end of the second year of the pandemic that number had dropped to only half.

What can be done? We recommend two approaches: (1) prioritise enduring relationships with patients and within teams; (2) push more decision authority to the physicians closest to patients.

Prioritise relationships with patients and within teams

The power of relationships is in our collective blind spot. Much clinically relevant knowledge is housed in the EHR and in the relationship itself. The workings of healthcare are advanced not only by electronic exchange of information but also by the trust, working knowledge and communication between healthcare workers speaking together about shared patients. By prioritising relationships at every level within the healthcare system, including culture, staffing, technology and policies, it is possible to decrease the high costs currently incurred from broken relationships between physicians and patients and among healthcare workers.

The first step is to reconceptualise healthcare away from the predominant transactional model, where healthcare is seen as a series of independent transactions that can be handled by any clinician, and instead recognising that at its core, healthcare is a relational enterprise.15 Care is safer, more efficient and more satisfying when the people involved have worked regularly together and thus know and trust each other. Continuity of relationships can be prioritised in patient scheduling, in team assignments and in communication across sites of service (eg, between home health staff and clinic staff).

EHRs also need to be implemented in ways that minimise work outside of work (also known as ‘pajama time’) and leave physicians personal time for relationships with family and friends. This means measuring and reducing the amount of time physicians spend on the EHR outside of work (currently 1–2 hours/day for US PCPs16 17) and developing policies and practices that ensure physicians are fully unplugged from their EHR when on vacation.

Relationships between physicians and their leaders also matter. Physicians who rank their immediate supervisor higher on attributes such as ‘is interested in my opinion’ and ‘encourages me to develop my skills and talents’ have a 66% lower odds of intending to leave their current practice.18 Feeling valued by one’s organisation reduces the odds of intending to reduce clinical work by 35% and of intending to leave by 60%.19

A move away from a physician-centric model of care to a team-centric model with strong staff support and task sharing allows the highest trained professionals to focus on the ‘solution shop’ work of solving unstructured problems and relationship building for which they are trained, and empower support staff to manage much of the standardised predictable work of the practice.20 It also means prioritising team stability so that the same support staff and physicians work together each day, allowing the efficiencies that come with team stability to flourish. This will require a substantial increase in the availability, affordability and sustainability of the allied health professional and nursing workforces.

Push more decision authority to the physicians closest to the patients

In the UK many diagnostic services, such as endoscopy, CT and MRI scanning, remain hospital-only services, denying GPs access to the tools they need to deliver better community-based care. Allowing practices to develop and deliver additional services at the local practice may well enhance the patient experience and give practices and GPs the opportunity to upgrade facilities and enhance the physical environment for staff and patients alike.

In the USA, this approach might involve moving away from the industrial model of central scheduling and instead colocate schedulers within the clinical teams. This would allow the scheduling template to be designed to more closely match the needs of the particular practice, whose clinicians may find that adjusting the start/stop times and appointment duration allows them to customise their schedule to the needs of their patients and staff. Schedulers who are a part of a specific care team can get to know their patients and their team’s work style. Embedded scheduling also allows joint problem solving between the clinical staff and the scheduler in service of the patient and more easily allows the team to accommodate patients’ urgent requests.


Burnout can be significantly reduced by prioritising and valuing relationships at every level within the healthcare system including culture, staffing technology and policies. By prioritising relationships and reducing burnout it is possible to decrease the high costs of physician turnover currently incurred when relationships between physicians, patients and team members are broken.

Ethics statements

Patient consent for publication



  • Twitter @Doc_Morrow

  • Contributors CS and JM contributed to the writing of this editorial.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Disclaimer Dr. Sinsky is employed by the American Medical Association. The opinions expressed in this article are those of the authors and should not be interpreted as American Medical Association policy.

  • Competing interests None declared.

  • Provenance and peer review Commissioned; internally peer reviewed.

Linked Articles