Abstract
Background The detection, assessment, and management of primary care poor performance raise difficult issues for all those involved. Guidance has largely focused on managing the most serious cases where patient safety is severely compromised. The management of primary care poor performance has become an increasingly important part of primary care trust (PCT) work, but its modes of presentation and prevalence are not well known.
Aim To report the prevalence, presentation modes, and management of primary care poor performance cases presenting to one PCT over a 5-year period.
Design of study A retrospective review of primary care poor performance cases in one district.
Setting St Helens PCT administered 35 practices with 130 GPs on the performers list, caring for 190 110 patients in North West England, UK.
Method Cases presenting during 2002–2007 were initially reviewed by the chair of the PCT clinical executive committee. Anonymised data were then jointly reviewed by the assessor and another experienced GP advisor.
Results There were 102 individual presentations (20 per year or one every 2–3 weeks) where clinician performance raised significant cause for concern occurred over the 5-year period. These concerns related to 37 individual clinicians, a range of 1–14 per clinician (mean 2.7). Whistleblowing by professional colleagues on 43 occasions was the most common presentation, of which 26 were from GPs about GPs. Patient complaints (18) were the second most common presentation. Twenty-seven clinicians were GPs, of whom the General Medical Council (GMC) were notified or involved in 13 cases. Clinicians were supported locally, and remedying was exclusively locally managed in 14 cases, and shared with an external organisation (such as the GMC or deanery) in another 12.
Conclusion Professional whistleblowing and patient complaints were the most common sources of presentation. Effective PCT teams are needed to manage clinicians whose performance gives cause for concern. Sufficient resources and both formal and informal ways of reporting concerns are essential.
INTRODUCTION
The detection, assessment, and management of primary care poor performance raise complex issues for all those involved. Difficult judgments must be made about current and future fitness to practise, and possible risks to patients and/or staff. These are aggravated by medicolegal concerns, the need for confidentiality, and maintaining trust between clinicians and those assessing and supporting them.
Since the publication of Good Medical Practice in 2001 by the General Medical Council (GMC),1 there has been greater awareness of governance, and the management of primary care poor performance has become a vital part of the work of all primary care organisations. Criminal cases inevitably have a high profile,2,3 yet the background work of assessing and supporting clinicians and teams who are struggling to perform adequately, or who risk patient safety, takes much more time.
Despite published guidance addressing some issues associated with primary care poor performance,4 it remains difficult to discern the magnitude of the problem, the typical ways in which it presents, and the outcomes that may be anticipated. A major review in 2006 concluded that: ‘The study of poor performance in healthcare is in its infancy’.5 This study reports a series of primary care poor performance cases from one district over a 5-year period: 2002–2007.
METHOD
St Helens Primary Care Trust (PCT) was responsible for primary care services, serving a population of 190 000 patients in an economically deprived industrial town in North-West England. During the 5-year study period there were, on average, 130 GPs (principals and non-principals) on its performers list (a register of accredited GPs allowed to perform primary care medical services), and 36 practices.
The St Helens Professional Performance Assessment Group was formed in 2002 and developed its Primary Care Poor Performance Assessment Framework based on experience with local issues, national reports and Good Medical Practice.1 Performance concerns were initially identified by complaints analysis, deviation from prescribing guidelines, and reports of problems by medicines management. Other markers such as the Quality and Outcomes Framework (QOF) were added later. The group regularly reported to the chief executive and PCT board. The framework was consistent with guidance from the National Clinical Assessment Service (formerly the National Clinical Assessment Authority),4 and had been reviewed by their regional advisor.
Risk was assessed according to threat to patient safety or probity, and classified as ‘red’ (requiring GMC or other professional body referral, or National Clinical Assessment Service direct involvement), ‘amber’ (requiring local remedying and support), or ‘green’ (isolated or resolved episode of minor concern). After initial informal evaluation, presentations deemed to be green cases were not taken on for further assessment within the performance-management process, but were recorded.
The PCT held records of the group's work over the 5-year period, which were reviewed by the chair of the PCT clinical executive committee and anonymised.
RESULTS
During the period 1 October 2002 to 30 September 2007, the professional performance assessment group managed 102 presentations in 37 clinicians where the level of performance gave significant, red, or amber cause for concern (range 1–13 presentations per clinician, mean 2.8). These comprised 27 GPs, five dentists, three nurses, and two community pharmacists. Table 1 demonstrates that multiple modes of presentation were common; for example, patient complaints and an episode of ill-health (self-referal). The annual incidence of new GP cases coming to the group's attention for red or amber concerns was 4.2 of 100 performers-list GPs (2.8 of 100 000 patients).
Table 1 Presentation source from 102 presentations, 37 clinicians.
A total of 27 GPs comprised 21 principals and six non-principals. Four of the non-principals were locums in the area. Of the remaining 23 doctors, 17 were from small practices (either one- or two-partner), and six from larger practices. The four locums all represented red-light risk; six red and 11 amber cases were from small practices, while three red and three amber were from large practices.
How this fits in
Clinical poor performance in primary care has become a priority for the public, professions, and Government, as recently highlighted by the work of the National Support Team for Tackling Health Inequalities. This study reports the incidence and prevalence over a 5-year period in one district. It shows that professional whistle-blowing can be the major source of presentation and confirms earlier work that professional misconduct and ill-health are the major presenting features of performance giving serious cause for concern. Successful management requires a major PCT commitment in time, expertise and money.
Table 2 shows how cases presented in terms of classification of performance issues according to type and risk to patients: 19/37 clinicians were classified as red-light risk, and 18/37 as amber. An average case had 1.9 presentational types (range 1–4), and presented 2.8 times (range 1–13).
Table 2 Classification of presentation by type of performance concern in 37 clinicians.
Six cases involved sexualised behaviour and six were alcohol related (one had both). Police disclosures within routine Criminal Records Bureau checks added information about three clinicians, including one new case being investigated.
Of the 27 GPs, management of seven affected their performers list status: one GP was removed from the performers list; one was suspended and later removed; one left general practice during the process; one remains on long-term ill-health; two GPs refusing appraisal received ‘28 day warning of removal from the performers list’; and a further one received a written warning regarding list status and future behaviour. One local GP referred for health and performance issues was also found not to be on any PCT performers list due to an administrative error. Excluding the first four, the other 23 GPs whose cases were managed by the group are known to be working as GPs today, all of whom are in unrestricted practice.
Table 3 indicates outside involvement and support in management of the 37 cases. Remedying work was undertaken solely within the PCT in 14 cases, and shared with another body, for example Mersey Deanery or GMC, in a further 12 cases. Eleven needed remedying or management outside the PCT.
Table 3 Organisational involvement beyond the PCT.
Work directly relating to the professional performance assessment group required significant time from both clinical and PCT support staff. Although the workload fluctuated considerably, the group needed several sessions of weekly input from a team including the medical director/chair of the clinical executive committee, GP tutor, the director of workforce (human resources), the governance lead/senior nurse, a senior PCT pharmacist, dental director, head of primary care, and local medical committee representation. Other senior colleagues were co-opted according to cases, for example local dental or pharmacist committee representation. The group also required a full-time administrative assistant, and data analyst time. Clinical assessors in this study of primary care poor performance cases were all current or past GP educationalists (GP tutors, GP trainers, and undergraduate teachers).
DISCUSSION
Summary of main findings
It is to be expected that GPs are the main professional group represented, as initially the research group was solely concerned with them. The most common presentational route was whistleblowing, particularly by GPs of GPs, who alone accounted for 26% of all presentations. Much work had been done in this district on whistleblowing processes, following a high-profile case of a local single-handed GP imprisoned for a series of sexual assaults on female patients.3 It is noteworthy that whistleblowing was over twice as common as patient complaints as a presentational source, despite the latter being traditionally the major source of concerns. Although a minority, 14% of cases of significant poor performance were of a possibly criminal nature, similar to the 11% reported by deaneries.6
The duration of cases from first presentation to closure was an average of 19 months (range 1–60 months). This appears lengthy, but is accounted for by the range of presentations (1–13), their management, and subsequent attempts at remedying and monitoring.
Cases with probity and health presentations were linked more closely to red risk in this study of primary care poor performance cases, substantially more so than those related to knowledge and skills. ‘Bad or sick, rather than poor’, has often been the impression of others working with high-risk cases. The National Clinical Assessment Service has reported that conduct accounted for two-thirds of referrals to them.7
Combinations of attitude, behaviour, and probity issues were seen in 84% of the more serious cases, reflecting some of the reactive type of presentations generated by patient complaints so often seen by PCTs. This is similar to a Tyneside series where less than one-quarter of cases had clinical care as a cause for concern.8 In contrast, more GPs assessed by the GMC had seriously deficient levels of knowledge (46% of cases) and similar levels of poor skills.9 In this study of primary care poor performance cases, the less serious locally-managed cases showed a much greater preponderance of knowledge, skills, or medicines-management issues. However, these criteria inevitably overlap and strict comparisons are hampered by the lack of strict definitions.
Strengths and limitations of the study
The study enabled calculation of the incidence and prevalence of performance issues among GPs, as all serious or potentially serious performance concerns have to be reported to the PCT. The series is large enough to show different modes of presentation and an indication of outcomes. However, it concerns one district and inevitably there will be local factors that would differ elsewhere. In particular, the study considers that whistleblowing will vary depending upon the trust people have in their PCT and its clinical leaders, and this has recently been given renewed emphasis.10 The study believes that early action was possible because of a constructive and supportive work climate,5 and that the local conviction of a criminal GP was the main factor in raising awareness among all of the clinicians and staff in the study.
Comparison with existing literature
There is little comparative published information in this area. Gray8 found 16 instances of poor performance among GPs in 2002–2003 (about 8% of 206 GPs in post in 2004).8 Referrals to the National Clinical Assessment Service in 2005–2007 were approximately 280 per annum (0.65% of all GPs each year). The service assessed 1075 GPs in 2001–2007, of whom 113 (11%) were suspended for some time.7 The GMC assessed 25 GPs under their performance procedures in the 3 years 1998–2000.9 It is noteworthy that a major review of why doctors perform poorly could not cite information on the incidence or prevalence of poor performance in doctors.5
Six cases included alcohol problems, of whom three were GPs (out of 130 on the performers list), which is therefore lower than other studies that have suggested 7% of doctors may be dependent on alcohol or other drugs.11
Implications for clinical practice and future research
It has been recognised that long-standing mediocre performance is often difficult to manage and remedy due to a lack of insight.5 In this study of primary care poor performance cases, the assessment of medical records from clinicians with chronic poor performance often featured reactive rather than proactive patient care patterns. Planned follow-up was rarely observed in clinicians' clinical records. Long-standing mediocre performance may be accompanied by low patient expectations leading to fewer complaints, which could be unrecognised by any PCT that relies excessively on complaints to trigger enquiries. This is an area that warrants future research.
PCTs have recently been substantially reorganised, and are probably struggling to handle performance issues.12 Informal contacts with other PCTs suggest many may have only looked at the worst poor performance, when GMC referral is very likely. PCTs may not acknowledge having poor performers, especially if they do not have any systematic or proactive methods of checking complaints against other markers of performance, or where there is no effective whistleblowing policy. An annual report of the numbers of cases of whistleblowing would be useful for any PCT board, and would say much about organisational maturity.
The cost for patients, practitioners, and the NHS makes continuing research on poor performance imperative. There is a need to know the demography of poor performance: how, why, and how quickly it has been detected. It is also important to know if it can be prevented or effectively remedied at an early stage, something that the GMC is now investigating.13 Outcomes will not be easy to measure, as complex assessment of sustained practitioner performance will be necessary — how many will still be causing concern 5 years later?
The combination of the publication of Good Medical Practice and the whistleblowing action by a local GP about another GP3 were the basis for this series. The study believes that it was the trust earned over time that enabled clinicians and others to come forward with their worries, which were then confirmed or refuted by wider enquiries. A robust, developmental but non-punitive approach could have led to this apparently high incidence of performance concerns, which is probably more reassuring than a misleadingly, artificially low incidence.
The management of poor performance in primary care requires high-level skills and experienced assessors within a complex field accompanied by significant PCT resources. This will link in future to both the ‘responsible officer’ and GMC affiliate roles. The study believes there is an overwhelming need in each district for informal opportunities to report concerns, combined with a system to record concerns that should function as ‘organisational memory’.14
Acknowledgments
We thank our colleagues in the primary care trust, local colleagues, practices, and teams for their cooperation.
Notes
Funding body
Not applicable
Ethical approval
Approved by the Chair of the responsible Ethics Committee on Merseyside (Sefton)
Competing interests
The authors have stated that there are none
- Received May 12, 2008.
- Revision received July 24, 2008.
- Accepted December 9, 2008.
- © British Journal of General Practice, 2009.
