Evidence Based Medicine - Introduction to Systematic Review

Evidence Based Medicine is an Approach of applying best evidences synthesized from scientific studies and their findings in clinical practice to take patient related decision. Since, Medicine is rapidly evolving phenomenon, there should be certain platforms which provide all the latest and most applicable findings all over the world in a concise and structured manner. With the adaptation of technology in Medical science and development of Research guidelines, Massive upsurge of Clinical Research took place, availability of treatment options  increases generating huge loads of information in different Platforms. There was difficulty for physicians to cope up new knowledge and Make an accurate decision considering literatures and patient populations.  After publication of  a book Named: "Effectiveness and Efficiency: Random Reflections on Health Services (1972)", by Scottish epidemiologist Archie Cochrane, Evidence Based Practice took approval in scientific communities and finally institutional approach was taken up from McMaster University research group led by David Sackett and Gordon Guyatt University. In 1992, the term "Evidence Based Medicine" was appeared for the first time in Medical literature. 

     

       Why Evidence Based Medicine is Necessary ?

• 

  For a physician in order to cope up with recent updates in the field, he/she has to go through minimal of 9 articles per day,  which sounds quite impossible considering the work load. Screening of single article written on the defined question of interest helps clinician to apply their Best possible treatment to Patient they are dealing with. Still, ultimate decision to choose the treatment option remains with physican based on his/her skill, efficiency of evidence and Patient choice. 
• Clinical Practice Guidelines are developed by different Institutions for specific clinical condition based upon rigorous systematic review and synthesis of  published medical
  

  
  

  literatures for specific diseases. Though these guidelines are  not protocol which must be followed but designed to recommend best possible treatment.  as for example: Guidelines for diagnosis and management of asthma, Management guidelines for stable ischemic heart disease, use of integrative therapy as supportive care in breast cancer patient, American college of rheumatology guidelines for the treatment of Rheumatoid arthritis and so on...
• Health policies are particularly developed based on Different Evidences available so far for treatment and Population of that particlar country and finally individual patient. This is referred specifically as Evidence Based Healthcare.  This type of healthcare is explicitely oriented to provide best available Cost effective treatment to the people of that particular country. 

       How Evidences are Graded In Medicine

Here simply using the term 'Evidence' raises one more question as what are the evidence we take into consideration ? which evidence is the strongest one ?  Following is the classification of different Evidence available in decreasing order  in terms of their strength while making decision. 

• Systematic review of Randomized Controlled trials by using Meta- Analysis. 
• Single or multiple Randomized Controlled Trials.
• quasi- randomized controlled trials
• Case studies and case series
• Observational studies (Cohort study, Case-Control study etc.)
• Case Reports and Case-series
• Experts Opinions

This system of hierarchy was developed by taking the Probability of Bias into considerations. In terms of Research design to generate the evidences, RCTs are considered to be the highest of all since they are systematically designed in such a way that chance of error and bias is minimum. On the contrary, though Experts opinions are always taken into consideration while making decision but by nature Personal Opinions and Judgement are always Biased in some way, hence considered as the poorest way to synthesize evidence. Again, systematic reviews of Randomized controlled trial by using Meta-analysis as statistical tool is in topmost rank of evidence pyramid. This model of ranking does not underestimate the efficacy of observational studies since, Meta-analysis does take care of quality of data than merely research design. Strength of evidence also depends upon the Nature of questions we ask, for example: if we want to evaluate the prognosis of a disease , Meta-analysis of Cohort study gives the solid result than RCT. Also, Pooly designed RCT are not better than adequately Conducted Cohort studies. 

                Introduction to Systematic Review  

Systematic review nowadays became synonym for Evidence Based Medicine, Specially after being endorsed by big platforms like Cochrane. Though there is no Consensus and standard definition of systematic review, Moher defined it as "Application of scientific strategies that limits bias to the systematic assembly, critical appraisal and synthesis of all relevant studies on a specific topic". According to Cochrane Library, systematic review attempts to identify, appraise and synthesize all the empirical evidence that meets pre-specified eligibility criteria to answer a specific research question. Clinical significance of an treatment can be determined by using various Statistically significant Individual trials on that specific disease/condition with the help of Systematic Review. 

As for example you want to know weather or not statin therapy can be beneficial over Red Yeast Rice in aged patient with hypercholesterolemia. While performing the systematic review this problem has been broken down in the following form as research questions

Patient/ Population: People Diagnosed with Hypercholesterolemia
Intervention: Red Yeast Rice
Control/Comparison: Standard Statin Therapy
Observation/ outcome: % decrease in Blood Cholesterol Level

      Major Characteristics of Systematic Reviews

Progress of Medical literature Review took place from traditional Narrative review to Contemporary Systematic Review by solving the fundamental limitations such as Selection bias by author, uncomprehensive Search strategies, Incomplete data synthesis and being descriptive rather than specific problem oriented. (https://researchsapling.blogspot.com/2020/03/doorstep-for-evidence-based-medicine.html)

•  Purpose of providing Reliable evidence on relevant clinical uncertainties addressing the focused PICOT question
•  Standardized 'Review protocol'  is followed finalized after peer review. Which gives the transparency and avoids conflicts of interest
• Always performed by Multiple authors, mostly by topics experts and Methodologist. 
• Reporting standards are always followed such as: PRISMA   ( Preferred Reporting items for systematic reviews and Meta-analysis), MOOSE ( Meta-analysis of observation studies in Epidemiology), MECIR ( methodological expectation of Cochrane intervention reviews) 
• Selection of studies based upon explicit inclusion/exclusion criteria with prespecified study designs and rationale of exclusion of study. 
• A rigid Search strategy is employed which includes Multiple database search for published and unpublished studies, ongoing trials from trial registries (https://clinicaltrials.gov) , with proper key and MESH terms without language restrictions. 
• Data synthesis methods are pre-defined for pooling effect estimates with 95% confidence interval, methods of heterogeneity, subgroup analysis and sensitivity analysis. 
• Review is always performed in multiple steps by multiple authors, results are interpreted on the basis of summery of findings and conclusions are made based upon the implications of policy and practice. 

Why does Meta-Analysis Come with Systematic review ?

Though these two systems are entirely different in terms of applicability, Systematic review is always performed using Meta-analysis as statistical tool. It is an statistical method to combine the result of several independent studies and to present the average result of them in  different graphical forms and also provides the greater statistical power than individual study. Since Meta-analysis is statistical tool to combine the result, it can be done with at least two studies and maximum of hundreds, the findings of Meta-analysis may not be useful or sometimes Misleading too without the proper context. Systematic Review gives the proper guidance to pool and present required data by filtering the unwanted loads of information. 

 Challenges and Limitations of Conducting Systematic Review

1.  In order to derive quality data and conclusions from systematic review, all the relevant trials should be searched and analysed which is the biggest challenge since hundreds of new research keeps happening yearly in the similar research questions. 
2.  There is a unavailability of skilled manpower who with expertise in subject matter as well as necessary technical skill to conduct a systematic Review. 
3.  There is high chance of missing scientifically Relevant study just because of astringent inclusions criteria and Statistical analysis employed. One study has found that, Dichotomous data was not being extracted in 8 citation out of 120 citations. 
4.  In context of Traditional Medical systems like Ayurveda, there are  extremely less number of Randomized controlled trial. Though there are number of research activities happening and generated data is found to be effective, due to meticulous regulatory and scientific guidelines, conducting large  scale research is being challenge always. In this context, Considering Highly specific models like Systematic review for evidence synthesis is discouraging for those practitioners and there is high chance of such Practices being Forsaken by the Healthcare regulation.  

                                                                                                    Thank You !!!

    


                        References

1. https://www.nccih.nih.gov/health/providers/clinicalpractice
2. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3789163/
3. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3124652/
4. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4975798/
5. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3024725/
6. https://www.cochranelibrary.com/cdsr/about-cdsr
7. https://www.cebm.net/wp-content/uploads/2014/10/Lecture-1-EBM-year4-introduction-2014.pdf
8. https://www.ncbi.nlm.nih.gov/pubmed/19469725