Unnecessary care in orthopaedic surgery
Abstract
Unnecessary care, where the potential for harm exceeds the potential for benefit, is widespread in medical care. Orthopaedic surgery is no exception. This has significant implications for patient safety and health care expenditure. This narrative review explores unnecessary care in orthopaedic surgery. There is wide geographic variation in orthopaedic surgical practice that cannot be explained by differences in local patient populations. Furthermore, many orthopaedic interventions lack adequate low-bias evidence to support their use. Quantifying the size of the problem is difficult, but the economic burden and morbidity associated with unnecessary care is likely to be significant. An evidence gap, evidence-practice gap, cognitive biases, and health system factors all contribute to unnecessary care in orthopaedic surgery. Unnecessary care is harming patients and incurring high costs. Solutions include increasing awareness of the problem, aligning financial incentives to high value care and away from low value care, and demanding low bias evidence where none exists.
Background
Unnecessary care, or overuse has been defined as the ‘provision of medical services for which the potential for harm exceeds the potential for benefit’.1 Brownlee et al. note that in considering unnecessary care, all interventions can be viewed to be on some point of a spectrum that stretches from tests and treatments that are universally beneficial when used on the appropriate patient, through to services that are ‘entirely ineffective, futile, or pose such a high risk of harm to all patients that they should never be delivered’.2
There is widespread recognition of the problem of overtreatment in medical care with initiatives to study and address the problem that include the Less is More series in JAMA,3 the Too Much Medicine initiative in The BMJ,4 Preventing Overdiagnosis conferences,5 the Wiser Healthcare research collaborative,6 and the international Choosing Wisely program.7
Across all fields of medicine, the ‘60-30-10 rule’ has been used to describe the burden of unnecessary care.8, 9 Studies in a variety of settings including the United States, Australia and the United Kingdom have shown approximately 60% of health care is of value and in keeping with guidelines, 30% is low-value, duplicative or unnecessary, and the remaining 10% is harmful.8
From an economic perspective, a 2013 report from the Institute of Medicine estimated “unnecessary services” increased health care spending by $750 billion annually in the United States (US) alone.10 Given that ‘low back and neck pain’ and ‘other musculoskeletal disorders’ are the conditions with the highest and second highest amount of associated health care spending in the US,11 the economic burden of unnecessary orthopaedic care is likely to be high.
Despite the potential patient harm and high costs, the area of overtreatment is not widely acknowledged or studied in surgical specialties. In this review, we focus specifically on the problem of overtreatment in musculoskeletal surgery.
Practice variation in orthopaedic surgery
Variation in the utilization of health services that cannot be explained by variation in patient illness or patient preferences is a well-known method of detecting low value2, 12 or preference-sensitive care.13
In orthopaedic surgery, large geographic variation has been found in spine surgery rates. In one study, the number of spine surgery procedures performed per capita was found to be more than five times higher in the US compared to in England and Scotland, and at least 40% higher in the US than in any of the 12 other developed countries studied.14 Significant variation has also been found within the United States. In a cross-sectional study using fee-for-service Medicare data from the Dartmouth Atlas Project, Weinstein et al. found that adjusted per capita lumbar fusion rates varied by a factor of more than 20 between regions.15 While another study reported preferential use of interbody fusion procedures for lumbar spine disorders in a fee-for-service compared to salaried remuneration model.16
Multiple studies have documented significant regional variations in knee arthroscopy rates, especially as the practice decreased (by up to 90% from its peak) following evidence revealing its limited utility in the context of degenerative change.17-21 Significant regional variation has also been reported in arthroscopic rotator cuff repair and subacromial decompression.22, 23 More recently, two studies have reported significant variation in health care utilization for elderly patients with hip fracture.24, 25
While some practice variation is appropriate,26 several factors suggest that a significant proportion of this variation is inappropriate and indicative of unnecessary care. First, a near linear relationship exists between the number of spine surgeons per capita and the number of spine surgery procedures performed13 and increased rates of rotator cuff repair and other orthopaedic procedures are seen in regions with higher orthopaedic surgeon supply.23 Second, higher rates of surgery are seen in the US where, in contrast to surgeons in the UK, surgeons have historically been paid under a fee-for-service model.27 In New South Wales, Australia (where public and private systems co-exist), rates of privately funded spinal surgery (which are more generously remunerated) are significantly higher and increasing more rapidly than those in the public sector.28 Finally, competition between hospitals appears to lead to increased intervention.29
Causes of overtreatment
Lack of supporting evidence
Across all fields of medicine, well conducted randomized control trials (RCTs) offer the least biased form of evidence to guide practice.30 A 2014 review article investigated the proportion of orthopaedic procedures supported by evidence from randomized controlled trials (RCTs) comparing operative intervention to non-operative alternatives. Just over half of all orthopaedic procedures performed had been compared to non-operative management in at least one RCT, 37% of procedures were supported by at least one RCT which showed surgery to be superior to a non-operative procedure, and only 20% were supported by at least one RCT that was considered low risk of bias.31 Concerningly, 16% of procedures had RCT evidence comparing operative and non-operative management which did not support the procedure over non-operative alternatives.
A 2021 study by Blom et al. studied the evidence base underlying 10 of the most common elective orthopaedic procedures by reviewing meta-analyses of RCTs (or in the absence of meta-analyses, other study designs) which compared interventions to non-operative management. They found RCT evidence supported the superiority of 2 of the 10 procedures over non-operative management, that 2 of the 10 procedures had no RCTs specifically comparing them with non-operative management, and that the evidence for the remaining 6 procedures showed no benefit of surgery compared to non-operative management. They commented that ‘no strong, high quality evidence base shows that many commonly performed elective orthopaedic procedures are more effective than non-operative alternatives’.32 Even amongst RCTs performed in orthopaedic surgery, few compare operative to non-operative management. A review focusing on the evidence for surgery for chronic musculoskeletal pain found that less than 1% of published RCTs on the most common surgical procedures for musculoskeletal pain compared surgery to not performing the procedure, and that most of those studies demonstrated no benefit of the procedure when compared to non-operative treatment.33
These data highlight two key problems relevant to unnecessary care in orthopaedic surgery: an evidence gap, where low bias evidence has not been generated, and an evidence-practice gap, where low bias evidence exists showing a lack of superiority over non-operative alternatives, yet the procedure is still commonly performed.
A higher proportion of medical interventions compared to surgical interventions are based on RCT evidence.34-37 Contributing to this is the discrepancy between the evidentiary threshold for approval of medicines versus the threshold for surgical procedures. Whereas drugs require low bias randomized trials against placebo or other drugs, the introduction and uptake of surgical procedures does not require this level of evidence. The surgical procedures that currently lack supporting evidence (including those that have evidence not supporting their use) were introduced in such a way, largely based on observational evidence alone.
Surgeon bias and social factors
A lack of evidence, and lack of understanding or awareness of evidence, can lead to uncertainty. Surgeons in general are trained to avoid uncertainty, and low levels of uncertainty have been shown in orthopaedic surgeons.38 Where there is uncertainty, clinicians may tend toward intervention due to biases that cause them to overestimate the effectiveness and underestimate the harms of their interventions. This has been shown in a systematic review of health care providers,39 and also for orthopaedic surgeons in particular, who have higher estimates of the outcomes of common orthopaedic procedures compared to the estimates provided by their patients.40, 41
Many orthopaedic conditions will improve with no treatment (natural history), or symptoms may regress to the mean over time or improve due to concomitant treatments (e.g., physiotherapy or anti-inflammatory medication). When orthopaedic surgeons see improvement in their patients following an intervention, they are at risk of assuming that the observed improvement is due to the intervention, falling victim to the post hoc ergo propter hoc fallacy – literally after this, therefore because of this.42
A number of cognitive biases including optimism bias (the tendency to overestimate the likelihood of a positive outcome compared to a negative outcome) and availability bias (the tendency to anchor decisions based on immediate examples that come to mind, for example, a recent happy patient) may also contribute to this overestimation of benefits.43
Of particular relevance in orthopaedic surgical practice is confirmation bias, the human tendency to ‘search for, interpret, and recall information in a way that confirms or supports one's prior beliefs or values’.44 A surgeon performing a large number of a particular procedure and who sees their patients improving, may be vulnerable to paying particular attention to and overestimating the quality of the evidence that supports their practice, and discounting evidence that doesn't.
Surgical training and culture may also play a role. During training, many common procedures are performed without questioning the evidence, and the social nature of surgical practice, whereby clinicians will tend to do what others surgeons are doing, means that practice can be guided by the beliefs of groups of physicians which are not necessarily evidence-based.45 Studies have shown that practice change amongst orthopaedic surgeons relies on cultural and social factors rather than evidence alone.46 This can be seen in the significant practice variation in spine surgery, where there is a lack of professional consensus and a poorer evidence-base compared to other subspecialties of orthopaedic surgery15, 47 as well as in the varying response to evidence against arthroscopic surgery for degenerative conditions in the knee, where some regions have shown large reductions in the rate of surgery and others have shown smaller rates of decrease or no decrease.19, 48-50
Patient demands and requests may also contribute to the burden of unnecessary care. While this is yet to be studied in orthopaedic surgeons, many primary care doctors report acquiescing to patient requests for unnecessary care51 and the expectations of orthopaedic patients vary greatly between countries.52
Health system issues
Health care payment models clearly influence practice and the fee-for-service model rewards volume and intensity.53, 54 In a survey of 2106 practicing US physicians, most respondents (70.8%) believed that doctors are more likely to perform unnecessary procedures when they profit from them.55 The Institute of Medicine Committee on Quality of Health Care has suggested better aligning the financial incentives with improvements in quality.56
Possible solutions
Possible solutions to minimize the burden of unnecessary care in orthopaedic surgery are outlined below and summarized in Table 1.
| Problem | Solution(s) |
|---|---|
| Evidence gap (a lack of high quality evidence). | Advocate for increased RCT evidence and higher evidentiary barriers for new interventions and existing procedures that lack such evidence. New procedures and technologies should be evaluated with effective scientific methods as part of their introduction into the market. |
| Evidence-practice gap (high quality evidence exists, but clinical practice has not responded appropriately). | Standardization of practice within and between institutions through accurate collection of practice variation data and structured feedback and discussion. Production and dissemination of evidence-based models of care and clinical practice guidelines. Increase surgeon and surgical resident education in biostatistics, epidemiology and the scientific method. |
| Surgeon and patient biases | Education of the public and surgeons around unnecessary care and the biases that may influence decision making. Effective implementation of shared decision making. |
| Health systems issues | Implementing payment models that align surgeon incentives with patient outcomes and reduce financial incentives for low value procedures. |
Evidence generation and incorporation
Generating more high quality evidence for orthopaedic interventions is essential for minimizing the burden of unnecessary care. It is apparent that much of modern orthopaedic surgical practice relies on historical and observational evidence and once a procedure becomes a standard of care, a large proportion of research is directed on how to best perform a procedure, but not whether performing the operation is better than not doing it.30 The generation of low bias evidence for new procedures will reduce the practice of introducing surgical procedures based on observational evidence alone, and only de-implementing the procedure once an RCT is done many years later. To avoid any stifling of innovation, new procedures could be evaluated as part of their introduction. This evaluation may be required by, or funded with input from regulatory bodies, insurers or professional bodies.
Education in the scientific method
Medical education, and especially surgical training, relies on an apprenticeship type model57 that tends to perpetuate existing practices. Appropriate education in the scientific method and basic research skills are essential yet often overlooked. A study of 277 medical residents across 11 residency programs in the United States found that most residents lacked the knowledge in biostatistics needed to interpret the results of published clinical research.58 This lack of scientific literacy and the apprenticeship model of training may be contributing to the evidence-practice gap. Surgical training and residency programs must therefore include more effective biostatistics and epidemiology in their curricula and attending surgeons should ensure they are equipped with the skillset to critically appraise the literature on which they base their practice. Finally, those with expertise and training in the scientific method must identify key messages and communicate these through effective systematic reviews or other syntheses of research findings.59
Standardization of practice
Sutherland et al. proposed an analytic framework for understanding and addressing unwarranted clinical variation in health care.60 They note that agency and motivation, evidence and judgement, and personal and organizational capacity all play a role and that measurement of variation is essential.60 In a systematic review of feedback approaches for unwarranted clinical variation, Harrison et al. note that “whilst methods for detecting variation are well-established, methods for determining variation that is unwarranted… are strongly debated.”61 Feedback approaches, therefore, are valuable to raise awareness of clinical variation and to stimulate discussion that can reveal when variation is justified.61 As such, accurate data collection regarding practice variation both within and between institutions, with structured feedback and discussion regarding these differences may help to address unnecessary care. From this, better standardization of practice, through the generation of evidence-based models of care and clinical practice guidelines, can occur.
Addressing biases and shared decision making
Educating the public about the potential harms and financial burden of overtreatment may be an effective step toward reducing unnecessary care. Equally, surgeons must be aware of unnecessary care and the biases that may influence their practice. Shared decision making (SDM) is also important. A scoping review by Niburski et al. found that SDM reduces surgical intervention rates and increases patient satisfaction.62
Addressing health systems and payment model issues
Aligning surgeon incentives with patient outcomes, a hallmark of value-based health care, may have a role in reducing unnecessary care. However, this must be done carefully due to the potential for unintended consequences. In the US, there has been a move away from fee-for-service models toward Alternative Payment Models such as the Bundled Payments for Care Improvement and Comprehensive Care for Joint Replacement models implemented by the Centers for Medicare and Medicaid Services.63 These initiatives have been shown to decrease per episode costs and high-cost care with no measurable change in outcomes.64 However, there is rising concern that these models may penalize patients with high health needs (due to ‘cherry-picking’ of patients) or penalize hospitals that serve such patients.65,66
Conclusion
Unnecessary care has large economic and patient safety implications. Although the amount of unnecessary care in orthopaedic surgery is difficult to quantify, geographic variation in practice, the link between financial incentives and volume of care, and the lack of high quality evidence mean that orthopaedic surgeons must be aware of this essential issue.
Solutions to minimize unnecessary care include generating and advocating for low bias evidence, better incorporation of the evidence into practice, optimizing payment models to align surgeon incentives with patient interests, and increasing education amongst surgeons and the public on biases, unnecessary care, and evidence based medicine.
Author contributions
Alex B. Boyle: Conceptualization; investigation; methodology; writing – original draft; writing – review and editing. Ian A. Harris: Conceptualization; investigation; methodology; supervision; writing – original draft; writing – review and editing.
Conflict of interest
None declared.
