Evidence-Based Medicine is NOT the Holy Grail
I am a Pulmonary/Critical Care physician who has worked as an intensivist in a large teaching hospital for the past 16 years. I have been responsible for the teaching rotations in the ICU for medical students, medical residents and pulmonary/critical care fellows throughout. I manage patients and teach daily. As a young trainee and physician in the 1980s-90s, access to relevant clinical data was limited. There was a bookcase of textbooks in the chief resident’s office, the library had some recent journals, citations found using Medline or references from texts that were more than a year or two old were searched on microfiche. In retrospect, I believe the effort necessary to unearth information was motivational and, essentially, forced diligence. I relied heavily on my knowledge of basic pathophysiology to explain bedside observations, and was able to be flexible with problem solving. In order to find the answer to the myriad questions I was faced with daily, I became a keen observer and avid questioner. Especially in the ICU, the need for near continuous manipulation of controllable variables to assess clinical response kept me alert, made me think and required multiple trips to the bedside.
Fortunate to have trained at Kings County, Bellevue and Massachusetts General Hospital, I was surrounded by like-minded people - we learned from each other. With information at a premium, sharing of knowledge and “running cases” by teachers and colleagues was mandatory. As information technology began to explode and the push for higher level evidence led to clinical research yielding information that, in theory, was more accurate and applicable, I could now be much more efficient and thorough. I evaluated and incorporated the steady influx of new data on “my terms”, building on a solid groundwork of basic science constructs. It quickly became clear that even the new “gold-standard” derived from multi-center, double-blinded, adequately powered, randomized controlled trials were far from immutable - often becoming obsolete after the next study. If my understanding of complex pathophysiologic concepts, by no means infallible, and the trial data were concordant, I was comfortable using the new information at the bedside. Rather than rubber-stamping the new EBM-derived principles, blindly “lifting’ them from journal to patient, I applied them when relevant, individually and fluidly, adapting my constructs and treatment based on my observations at the bedside. Over time, as my understanding of basic concepts was reinforced, added onto or torn down and rebuilt and I became aware of the limitations and capricious nature of EBM, I continued to grow as a clinician.
In the wrong hands, knowledge and utilization of EBM is potentially dangerous. In my experience, our newest doctors are often taught about the literature “in a vacuum”. Medical societies rapidly endorse intricate, multi-level algorithms and recommendations based on new perceived high-quality studies without much guidance on practical application. In light of the necessity to modify essentially all treatments based on innumerable individual differences and continuously reassess and adapt based on expected time-dependent variations, an EBM “mandate” may be completely nonsensical at best to harmful at worst. In addition, the recommendations are held in such high esteem that they are often stubbornly adhered to despite the clear “physiologic objections” of the patient.
This is bad medicine!
Arguably, the two most important positive trials in critical care over the past 20 years have been the “low tidal volume” ventilation trial by the ARDSnet investigators and the Early Goal-Directed Therapy trial [1, 2]. In the ARDSnet trial, with the rationale of avoiding alveolar over-distention in the setting of acute respiratory distress syndrome (ARDS), patients were randomized to 6 cc/kg ideal body weight or 12 cc/kg ideal body weight. There was a significant mortality benefit using lower tidal volumes. While I believe that avoiding alveolar hyperinflation is extremely important in severe diffuse lung injury and using the lower volume regimen prescribed was superior to the classic “high stress” ventilation used in these patients, the utilization of a discrete variable (e.g., 6 cc/kg) in lung injury with a “normal” distribution of severity is ludicrous. In fact, the average pre-trial tidal volume utilized was 9 cc/kg. It’s unclear if the results were related to randomization of one group to a lung-protective strategy or the other group to a known injurious approach. More importantly, though 6 cc/kg may very well be better than 12 cc/kg, on average, no patient is “average”. There is no question that a safe tidal volume depends on the degree of lung damage and the number of open alveoli, the latter influenced by a myriad of patient related and non-patient related variables. In other words, sometimes using 12 cc/kg might not be excessive and sometimes 6 cc/kg could cause lung injury.
In the second paper mentioned, patients in early septic shock were randomized to a targeted multivariable resuscitation algorithm (fluid, blood, dobutamine) in order to rapidly restore normal tissue oxygenation (through normalization of central venous oxygenation) or to very reasonable “standard of care”. The goal-directed arm of the protocol, which required placement of an upper body central venous catheter, had significantly lower mortality. The algorithm was initially promoted in toto in theinternational guidelines for treatment of septic shock [3], which I believe, was surprisingly misguided. Not only is it not clear (or studied) which aspect/s of the algorithm contributed to the outcome, but it’s quite possible that elements of the overall “prescription” were harmful. Again, targeting and achieving a discrete numerical result as an endpoint for resuscitation in a heterogeneous group and improving overall survival is far from proving that this strategy is optimal for YOUR patient. Interestingly, a more recent trial, ProCESS [4], showed equivalence between protocol-driven and standard therapy!
EBM-derived recommendations and guidelines are often quickly supported by our medical societies, facilitating utilization and likely improving "average" care. In isolation, however, they are much less relevant to an individual patient. We should bestriving for “best” care. On the other hand, do not "throw out the baby with the bathwater". We should aim to be evidence-based. The caveat is the evidence needs to begin with sound physiology, be supported by studies of the highest caliber available and related treatment needs to be skillfully adapted to the single patient. Ongoing interventions must account for observations made at the bedside as well as newly available data. Unfortunately, as a teacher, I am struck by the current unyielding reliance on often capricious study results. The knowledge base of the typical health care practitioner and time spent at the bedside has dwindled. Thenumerous adjustments necessary to intelligently apply specific EBM-derived treatment plan or algorithm to a distinct patient are immensely under-appreciated. I believe that the superficial knowledge and misguided use of the instantaneously available guidelines from even our EBM “home-runs” has had a negative effect on modern healthcare delivery. Unfortunately, this phenomenon is also directly related to a patient’s severity of illness, magnifying its impact.
A carpenter is only as good as his tools, but no matter how good his tools are….
REFERENCES
The Acute Respiratory Distress Syndrome Network. Ventilation with Lower Tidal Volumes as Compared with Traditional Tidal Volumes for Acute Lung Injury and the Acute Respiratory Distress Syndrome. N Engl J Med. 2000; 342: 1301-1308
E Rivers, B Nguyen, S Havstad, J Ressler, A Muzzin, B Knoblic et al. for the Early Goal-Directed Therapy Collaborative Group. Early Goal-Directed Therapy in the Treatment of Severe Sepsis and Septic Shock. N Engl J Med 2001; 345: 1368-137.
Dellinger R, Carlet J, Masur H, Gerlach H, Calandra T, Cohen J et al. Surviving Sepsis Campaign Management Guidelines Committee. Surviving Sepsis Campaign guidelines for management of severe sepsis and septic shock. Intensive Care Med 30: 536-55
The ProCESS Investigators. A Randomized Trial of Protocol-Based Care for Early Septic Shock. N Engl J Med. 2014; 370: 1683-1693