Evidence-based Medicine and Data Sharing in Outpatient Plastic Surgery

The Authors discuss the importance of evidence based medicine (EBM) as related particularly to outpatient surgery. They discuss the five core steps in current EBM and take the reader through each of the steps, listing methods the surgeon can follow to achieve a thorough and relevant evidence based plan. Challenges of EBM such as reporting bias and evidence grading are discussed along with solutions and tools to meet those challenges. The article concludes with a look at data sharing as a means of enabling surgeons to access outcomes and specific aspects of care for a surgical procedure.

Key points

•

Plastic surgery organizations have put forth initiatives to improve EBM skills among plastic surgeons.
•

Small steps toward understanding and mastering the practice increase the level of expertise, improve outcomes for patients, and raise the bar for patient safety.
•

Openness toward data sharing and better standards for implementing it strengthen the evidence base and lead to better health care quality and optimal patient outcomes.
•

Modern EBM is composed of five core steps: (1) assessing clinical practice to identify an important patient or policy problem; (2) asking clinical questions that are related to the problem and constructed to facilitate a sufficient literature search; (3) acquiring the best available evidence to answer the clinical question; (4) appraising the validity, importance, and clinical use of the evidence; and (5) applying evidence that is relevant to individual patients and aligned with their preferences and values.

Introduction

To become a plastic and reconstructive surgeon requires years of graduated responsibility in a structured residency training program. During this training period prospective surgeons gain technical expertise to perform surgical procedures and manage patient care through careful observation of their mentors and increasing responsibility, much like an apprenticeship. Although the mentors are incredibly skilled and experts in their field, relying on expert opinion to make treatment decisions is no longer sufficient in the realm of evidence-based medicine (EBM). As pressures from regulators and payers are increasing, the performance of surgeons is being scrutinized like never before. Treatment decisions that were once based on various forms of evidence, such as years of surgical practice, successes with previous patients, and information from the surgical literature, must now be supported by strong clinical evidence to be considered acceptable by the wider health care community.

Plastic surgeons must be dedicated to patient safety and quality improvement in all areas of practice. The use of EBM is particularly important for outpatient surgery, because approximately 80% of surgeries in the United States are performed as outpatient procedures and many ambulatory surgery facilities are unaccredited and in many states uninspected, with no regulatory oversight. Efforts are underway to promote accreditation of ambulatory surgery centers, but learning EBM and implementing its principles are also critical for improving quality and patient outcomes in the outpatient setting.

History and principles of EBM

EBM is the conscientious, explicit, and judicious use of current best evidence, combined with individual clinical expertise and patient preferences and values, in making decisions about the care of individual patients. Rudimentary accounts of evidence-based practice date back to ancient times ; however, the term “evidence-based medicine” did not exist until the early 1990s, when it was first published in the ACP Journal Club and later introduced to the wider medical community by the Evidence-Based Medicine Working Group. Initially, the concept of EBM was met with much criticism, because it incorrectly implied that the practice of medicine was unscientific. Over time, health care professionals began to understand that EBM was a framework and cultural standard for finding and applying the best evidence to guide treatment decisions.

Although the acceptance and practice of EBM has increased since the early 1990s, audits of medical and surgical procedures have revealed that low levels of evidence are still guiding treatment decisions. Importantly, these audits were conducted at single institutions shortly after EBM emerged, so it is unclear if the findings are representative of most health care facilities today. Nevertheless, there are always be cases for which little to no evidence is available, and clinicians need to rely on their best judgment and best available evidence at the time, but more work is needed to increase awareness of EBM and promote better research practices to enhance the evidence base and ensure that most treatment decisions are based on sound evidence.

Unfortunately, even when evidence is available, the research findings have the potential to be biased. This bias, or systematic error, is a reproducible error in study design or conduct that leads to systematic deviations from the underlying truth. Basing treatment decisions on biased information or inadequately tested theories can have devastating effects on patient outcomes. Therefore, clinicians need an understanding of EBM principles to help identify the best evidence to guide practice.

Modern EBM is composed of five core steps: (1) assessing clinical practice to identify an important patient or policy problem; (2) asking clinical questions that are related to the problem and constructed to facilitate a sufficient literature search; (3) acquiring the best available evidence to answer the clinical question; (4) appraising the validity, importance, and clinical use of the evidence; and (5) applying evidence that is relevant to individual patients and aligned with their preferences and values. Table 1 provides an overview of these steps and helpful methods for accomplishing each step, which are also described herein.

Table 1

Core steps of EBM and methods for performing each step

Step	Description	Methods
Assess	Recognize, classify, and prioritize important patient or policy problems	Assess problems in individual practice Search review articles and clinical practice guidelines to identify unmet medical needs Discuss areas of interest and potential clinical issues with colleagues Listen to patients to identify unmet needs that are important to patients
Ask	Construct clinical questions that facilitate an efficient search for evidence	Use PICO to develop good clinical questions: identify the patient/population/problem, intervention, comparison, outcomes
Acquire	Gather important and convincing evidence from high-quality repositories of the health literature	Use STARTLITE to develop a search strategy. Identify the: Sampling strategy: all or selected studies Type of study: systematic reviews, RCT, and so forth Approaches: electronic search, hand search, and so forth Limits: English-language articles, humans, age of patients Inclusion and exclusion: criteria for including or excluding studies Terms: search terms (MeSH terms, key words, and so forth) Electronic sources: electronic databases (eg, MEDLINE, CINAHL, Cochrane Library, and so forth)
Appraise	Systematically check best available evidence for indications of validity, importance, and usefulness	Use critical appraisal tools and resources to assess for potential biases: Center for Evidence Based Medicine, http://www.cebm.net/ Users’ Guides to the Medical Literature (JAMAevidence), http://jamaevidence.com/ Critical Appraisal Skills Program, http://www.casp-uk.net/ Grading of Recommendations Assessment Development and Evaluation Working Group http://www.gradeworkinggroup.org/index.htm
Apply	Interpret the applicability of evidence to specific problems, given patient preferences and values	Weigh the risks and benefits of the treatment option for each patient Ensure that the treatment option aligns with the patient’s values and preferences Develop plans for implementing the evidence in private practice and larger health care facilities (knowledge translation)

After a problem has been identified, developing a good clinical question facilitates a successful literature search. Clinical questions can be about treatment; harm; prognosis; diagnosis; or cost-effectiveness (economic analysis). The PICO ( P opulation, I ntervention, C omparison, O utcome) method is commonly used to develop clinical questions. An answerable clinical question in outpatient plastic surgery may be: “For women with breast hypertrophy, does breast reduction compared with physical therapy result in better health-related quality of life?,” where the patient population is women with breast hypertrophy, the intervention is breast reduction, the comparison intervention is physical therapy, and the outcome is health-related quality of life.

Finding the evidence to answer the clinical question involves several steps. First is to define the literature search strategy. The STARLITE ( S ampling strategy, T ype of study, A pproaches, L imits, I nclusion/exclusion criteria, T erms, and E lectronic sources) method is a useful tool for developing a search strategy. The type of clinical question helps to narrow the search to specific types of studies. For example, clinical questions about therapy are best answered with data from randomized controlled trials (RCTs), whereas clinical questions about prognosis are best answered with data from cohort designs. All types of study designs can be included in the search to optimize results, especially when little evidence exists for a particular question or a particular study design is not feasible or ethical; however, searches should aim to identify studies with the highest levels of evidence to best inform clinical decisions. Searching several bibliographic databases, including repositories of gray literature, and hand searching the bibliographies of relevant articles increases the likelihood that the body of evidence that has been collected is comprehensive and represents the underlying truth.

A common misconception in EBM is that the study design alone determines the strength of the evidence. Although RCTs can provide strong evidence, they are not created equal. The results of small, poorly designed RCTs can be misleading. Therefore, all types of studies should be appraised to determine their validity, importance, and clinical applicability. Critical appraisal is the process by which the methodologic quality of a study is screened for potential biases. Several critical appraisal tools have been developed to aid clinicians with this process. Importantly, each type of study is evaluated by a specific set of criteria. Box 1 provides an example of a critical appraisal tool for evaluating an RCT.

Box 1

Assessment for Selection Bias
•

Where patients recruited appropriately?
•

Was allocation concealed?
•

Were participants randomized appropriately?
•

Were treatment groups similar with respect to known and unknown prognostic factors?
•

Were confounders addressed?
•

Were data complete for at least 80% of participants in each group?
•

Were any significant differences found between participants who were lost to follow-up and those who completed follow-up?
•

Were participants analyzed in the group to which they were randomized (intention-to-treat)?

Assessment for Intervention Bias
•

Was the intervention well described?
•

Was the intervention implemented similarly in all participants (ie, could level of surgeon expertise influence how the procedure was performed; were there any protocol deviations)?
•

Was the caregiver (eg, surgeon) masked?

Assessment for Measurement Bias
•

Were the participants, outcome assessors, and data analysts masked?
•

Were outcomes measured similarly and with valid, defined criteria?
•

Was follow-up sufficient to detect all outcomes of interest?

Assessment for Type II Error
•

Was power sufficient to detect differences for each measured outcome?

Critical appraisal tool for an RCT

If after critical appraisal the study is deemed to be of high quality for the particular study design, it is then assigned a level of evidence according to the clinical question that the study attempted to answer. Numerous rating scales and their iterations have been published over the years; many are based on the first rating scale that was published by the Canadian Task Force and later refined by Sackett and the Center for Evidence Based Medicine. Typically, levels of evidence range from I to V, with I representing the highest level or strongest evidence and V representing the lowest level or weakest evidence. Plastic and Reconstructive Surgery has implemented a pyramid system to identify the clinical question and level of evidence of studies published in the journal ( Fig. 1 ). The pyramid is located on the first page of each article, providing a prominent visual cue that alerts the reader to the strength of the evidence provided by the study.

After critically appraising each study for a particular clinical question, the collective body of evidence is graded according to its strength in guiding a clinical recommendation. Like rating scales for levels of evidence, a variety of grading scales are available for recommendations. Recommendations are typically graded from A to D, with A representing the strongest recommendation and D representing the weakest recommendation. High levels of evidence often lead to strong clinical recommendations; however, this is not always the case. For example, high-level evidence suggests that continuous anticoagulation therapy reduces the risk of recurrent thrombosis in patients who have had an unprovoked deep vein thrombosis. However, continuous treatment with an anticoagulant also increases the risk of bleeding and is inconvenient for the patient. Therefore, weighing the benefits and risks of continuous anticoagulation therapy for this patient population may result in only a weak to moderate recommendation. Tables 2 and 3 illustrate the evidence and recommendation scales used by the American Society of Plastic Surgeons and the American Society of Aesthetic Plastic Surgery.

Table 2

Scale for rating the level of evidence of therapeutic studies ^a

Level of Evidence	Qualifying Studies
I	High-quality, multicentered or single-centered, randomized controlled trial with adequate power; or systematic review of these studies
II	Lesser-quality, randomized controlled trial; prospective cohort or comparative study; or systematic review of these studies
III	Retrospective cohort or comparative study, case-control study, or systematic review of these studies
IV	Case series with pretest and posttest, or only posttest
V	Expert opinion developed by consensus process; case report or clinical example; or evidence based on physiology, bench research, or “first principles”