A survey of more than 6,000 US physicians shows an association between burnout rates and doctors’ use of electronic record systems.
Declaration of interest: I use an electronic health record system at work, and it frequently makes me want to throw my computer out of the window…
Burnout is sadly common amongst healthcare workers, and a cause of significant mobility and mortality. Rates of depressive disorder are increased in physicians, and more worrisome still, deaths by suicide are higher than in the general population: 1.4x higher in male and 2.3x higher in female doctors, according to a study some years back (Schernhammer, 2004).
The condition itself was first described by Freudenberg in 1973 and several frameworks for its defining features have been put forward. Burnout now merits an entry in ICD-10 under Z73: Problems related to life-management difficulty. Subsequent work by social psychologists Christina Maslach and Susan Jackson helped to enshrine its 3 key domains of emotional exhaustion, depersonalisation and reduced personal achievement, and led to the development of the Maslach Burnout Inventory (MBI). This well validated self-report 22 item scale can be used by people to assess whether they are at risk of burnout. The scale itself is copyrighted and available from MindGarden.com but its components are detailed in the original paper by Maslach et al in the Journal of Occupational Behaviour.
Workers who are feeling jaded and exhausted are more likely to be inefficient, make greater numbers of errors and have higher rates of illness. Some will leave the profession altogether. Clearly it is in the interests of patients therefore to have their doctors working in an unburnt out state. Hence since identifying the problem, many studies have looked at both its extent and prevalence, and the possible associated factors and causes. Check out the good summaries from Sharmila Devi in The Lancet and Shailesh Kumar in World Psychiatry and Healthcare. The aim of all of this being to ameliorate the effects in individuals, and in designing systems to better protect workers against burning out.
The last decade has seen an explosion in the use of electronic health records (eHRs) and other computerised tools for ordering and checking investigations and results, coding diagnoses, dictating and signing off discharge letters, and interacting directly with patients through electronic patient portals. Many organisations have favoured these over traditional handwritten records for ease of sharing of information, legibility, and integrations with other databases to build expert systems which improve patient safety; for example if you try to prescribe contraindicated medications, the system will alert you with a warning.
On the flip side though, systems can be a time-consuming hinderance, leading to increased administrative work needing to be completed by the physicians themselves. Poorly designed, clunky and poorly linked systems can lead to wasted time spent staring at a computer screen rather than being with patients; all of which surely leads to a less satisfying job?
One angle on all of this is the topic of the main paper for this blog; a survey of US physicians by Shanafelt and colleagues (Shanafelt et al, 2016). They report on the associations between burnout in doctors and the growing use of electronic recording systems in healthcare.
The authors note that there are many reasons to think that eHRs are a culprit for the high rates of burnout seen, and that earlier studies were small and conducted in the early 2000’s when such systems were uncommon and in their infancy. They conducted a survey of 35,922 physicians taken as a representative sample of the 835,451 doctors on the American Medical Association Masterfile. The same authors have used this sample set in a previous study looking at changes in burnout and satisfaction with work-life balance in doctors more generally; which alarmingly found that over half of US physicians were experiencing burnout.
For this study, of 6,880 who responded (19.2%), 6,560 of these in active clinical practice were used as the dataset. Questions about the demographics of the respondents were asked, along with asking them to complete the MBI. Those that rated high on the depersonalisation or emotional exhaustion sub-scales were counted as having an episode of burnout. Information was also gathered about the “electronic environment” and clerical duties of the doctors, looking at:
- Use of eHRs, Computerised Physician Order Entry (CPOE – basically ordering blood tests on a computer) and patient portals
- Methods used to document clinical work
- Opinions about effects of electronic systems on quality of care and efficiency (5 point Likert scale from strongly agree to strongly disagree)
- Satisfaction with administrative tasks directly and indirectly related to clinical care of patients (5 point Likert scale from strongly agree to strongly disagree)
The results were analysed using Kruskal-Wallis tests or chi-squared tests to compare associations between variables, whilst logistic regression was used to identify factors significantly contributing to differences in burnout and satisfaction scores.
The demographics did not differ hugely from the US AMA Masterfile norms, except that the median age of these doctors was older (56 vs 51.5 yrs). Two thirds were male, three quarters worked outside primary care services, with just over half (55.2%) in private practice. Respondents covered a wide range of specialties, including psychiatry. They worked a median of 50 hours a week, with a median of one night on call per week.
Burnout ratings revealed that 55.3% (3,586 doctors) were burned out; i.e. scoring high on emotional exhaustion or depersonalisation sub scales of the MBI.
Use of e-health systems
When it came to use of technologies and administration, 84.5% of doctors were using an eHR, but only 36.1% were satisfied or highly satisfied with them. It also appeared that different specialties rated their satisfaction differently, with about half of pathologists, paediatricians and ob/gyn doctors happy with them, but only a quarter of ophthalmologists, occupational medics and otolaryngologists being satisfied with theirs. A similar amount (36.3%) agreed or strongly agreed that they had improved patient care, and only 23% agreed or strongly agreed that they had improved the doctors’ efficiency.
For CPOEs, 76% were using them, but only 38.1% were satisfied or very satisfied with them. Patient portals were used only by 26.1% of the sample, and again of these only 35.2% and 22% agreed or strongly agreed that these improved care or efficiency respectively.
On the questions relating to clerical tasks the majority (61.6%) of doctors were having to handwrite or type their own notes, while about a sixth each used self-entered voice recognition software or a dictation service. 284 lucky individuals (4.5%) had their notes entered by someone else acting as a scribe for them! Only 37.3% agreed or strongly agreed that the amount of time that they spent on admin tasks was reasonable, and only 25.6% felt that the time spent on clerical tasks not directly related to patient care was reasonable. Again there were differences between specialties – this time with pathologists and radiologists being most satisfied with their clerical burden.
Use of e-health systems associated with burnout
When the relationships between these variables and burnout was factored in, the authors found that doctors using eHRs, CPOE and patient portals were less satisfied with their direct and indirect admin role. Those using eHRs and CPOE had significantly higher rates of burnout to those who did not (57.2% vs 44.6% p<0.001), and interestingly use of voice recognition software was also associated with dissatisfaction with clerical burden and burnout compared to other methods of data entry. After accounting for demographic factors, use of eHRs or CPOEs was associated with dissatisfaction with direct clinical admin time, and use of a CPOE system was associated with higher rates of burnout.
Shanafelt and his colleagues show that although now widely used, eHRs and their ilk are poorly regarded by the doctors using them. There were significant levels of unhappiness with the added time burden of using these systems, and significant disagreement as to whether they actually improved patient care at all. Doctors largely feel that they are not happy with the amount of time spent on these tasks, and felt that they had worsened their efficiency. It is perhaps unsurprising to find that burnout rates were higher in doctors in this situation.
The authors go on to ponder how (or whether) such systems can be better used, for example by finding out which of the systems contribute least to dissatisfaction or inefficiency, or by developing systems that do not add to the clerical burden of clinical staff. One approach they suggest is the use of scribes to order tests or document clinical encounters, thus freeing up the physician’s time to spend interacting with patients. Another model is the use of advanced care teams with non-physicians being trained to assist with order entry, care co-ordination, responding to patient portal and email communications, etc.
In the interim, things may improve on their own, as the paper comments that younger clinicians (those <40) were more satisfied with the electronic systems, albeit with a majority still unhappy with the effects on admin task burden and still just as burnt out.
Some of the limitations of the present study include a low response rate, which given the electronic nature of the survey may reflect those most skilled with computing technology. The authors did not ask about the type of systems used, to be able to see if any particularly good or bad ones existed, and nor did they look at patient satisfaction for clinical interactions where the doctor was using a computerised system with them. All such things would be good topics for future studies to investigate.
Perhaps confirming my own anecdotal frustrations surrounding eHRs, this large study of American physicians finds a significant association between use of electronic record systems and reduced efficiency, greater administrative burden and higher rates of burnout. This suggests a couple of things:
- There is hope that the next generation of these systems to come along will be better designed, so as to fit seamlessly into clinical work
- There is probably a whole lot of money to be made for anyone with the clinical and IT skills to design such a system!
Shanafelt, Tait D. et al. (2016) Relationship Between Clerical Burden and Characteristics of the Electronic Environment With Physician Burnout and Professional Satisfaction. Mayo Clinic Proceedings , Vol91(7): 836 – 848 [Abstract]
Maslach C. and Jackson S. E. (1981) The measurement of experienced burnout (PDF). Journal of Occupational Behaviour, Vol. 2: 99-113
Devi S. (2011) Doctors in Distress (PDF). The Lancet Vol. 377 February 5, 2011: 454-455
Kumar S. (2007) Burnout in Psychiatrists. World Psychiatry. Vol. 6(3): 186–189.
Kumar S. (2016) Burnout and Doctors: Prevalence, Prevention and Intervention. Healthcare 2016, Vol. 4(3): 37
Schernhammer ES, Colditz GA. (2004) Suicide rates among physicians: a quantitative and gender assessment (meta-analysis). Am J Psychiatry. 2004 Dec;161(12):2295-302. [Abstract]
Shanafelt T. D., et al (2015) Changes in Burnout and Satisfaction With Work-Life Balance in Physicians and the General US Working Population Between 2011 and 2014. Mayo Clinic Proceedings, Vol. 90(12): 1600-13 [Abstract]