-By Dr. Akash G Prabhune
A epidemiological study can be broadly classified as observational or experimental. Observational studies can be further classified as Descriptive and Analytical. Descriptive studies are usually small clinical observation like case series, case reports which are an effort to scientifically present a clinical scenario(Pearce, 2012). Analytical studies can be classified into Cross Sectional, Case Control, and Cohort studies depending on the conduct of study. A cross sectional study is like a snapshot of time while a case control study is where events that have happened in past are to be related with current set of events. A cohort on other hand is when we follow a group of individuals for a given period. Experimental designs are the ones wherein study investigators test out the effect of an intervention (the intervention can be a new drug, new vaccine, new exercise, or a new surgical procedure). The experimental designs are further classified into randomized controlled trials (RCT), and quasi experimental trials. The difference between the two is one randomly allocates participant to either of the trial group while other omits the randomization part.(Pearce, 2012) The Flow Chart 1. (Pearce, 2012)gives a brief idea on
classification of epidemiological studies. In this article, we will dig deeper into randomized controlled trails (RCT) or what is commonly referred as clinical trials.
When we look at the ability of the study design to prove the evidence then RCT’s are at the apex of the research pyramid because of the randomization which inherently takes care of known and unknown confounders(Moher et al., 2010).RCTs are rigorously conducted studies which have a set of schools when it comes statistical analysis. A statistical analysis of an RCT can be conducted as “Intention to treat” (IIT) or “per protocol” (PP). IIT is a strategy wherein data on all the participants who were enrolled in the trail after random allocation must be analysed regardless of their status (cured, death, drop out, loss to follow-up) at the end of the trail. PP is a strategy wherein the data form non-complaint (drop outs, loss to follow-up) participants is not analysed.(Wassertheil-Smoller and Kim, 2010) For an example refer to Flow chart 2.
In the fluconazole trial, a total of 700 participants were randomly allocated to either intervention arm with fluconazole or control arm with placebo. As the trial progressed 10 people did not receive treatment in intervention arm and 15 in control arm, 10 people were lost to follow up in intervention arm and 7 in control arm, 5 people discontinued medication in intervention arm while 4 in control arm.Finally, at the end of the trial if we were to analyse the results using ITT strategy we must analyse data on all the 350 randomly allocated participants, even though they were non-complaint. If we were to analyse the trial data using PP strategy, we must analyse 325 participants from intervention arm and 324 from control arm as we need to eliminate non-compliant participants. The advantage of ITT is preserves the baseline integrity of random allocation and guards against drop out bias and it gives a real-life data on performance of intervention giving idea on adherence to treatment regimen. PP gives idea about the efficacy and excluding non-complaint participants ensures results are not skewed due to large number of loss to follow-ups or drop outs.(Gupta, 2011)Consolidated Standards of Reporting Trials (CONSORT) group which works on reporting standards for RCTs clearly states that in case of a drug trial or vaccine trial ITT strategy should be followed as it gives results which share resemblance to real life scenarios(Schulz, 2010).
Common concepts to be applied for analysing RCT data(Moher et al., 2010)
- Understand the type of data – The data collected for a RCT or in case any epidemiological study can be broadly classified into categorical (Age Groups, Gender) and Continuous (Height, Weight). The understanding of data type is necessary because categorical data needs to be analysed differently and continuous data requires its different set of analytical tests.
- Understand the dependent (Outcome Variable) and independent variable (Exposure Variable) – Dependent variable is outcome of the intervention applied. Independent variable can be anything from risk factor to a diseases (smoking), a condition (Hyperglycaemia), a disease (Malaria) which is goanna effect our outcome.
- Understand the relation between the dependent and independent variable – There are instances when a dependent variable is categorical and independent is continuous, the type of statistical tests to be applied varies due to the type and relationship between these two players. Table 1. Summarizes the statistical tests to be applied to various permutations and combinations of dependent and independent variables.
- Understand the statistical software’s – Statistical software packages are available in abundance in market which can analyse the data, howsoever Table 2. Lists the best compatible software’s based on the study type.
|Table 2. Statistical Packages and Study designs(Dembe et al., 2011)|
|Study Type||Analysis Type||Best suited statistical package|
|Cross Sectional||Descriptive||Epi Data|
|Cross Sectional||Analytical||Epi Data, SPSS|
|Case Control||Descriptive, Analytical||SPSS, SAS, R, STATA|
|Cohort||Descriptive, Analytical||SPSS, SAS, R, STATA|
|Randomized Controlled Trials||Descriptive, Analytical||SPSS, STATA, SAS|
|Quasi Experimental||Descriptive, Analytical||SPSS, STATA, SAS|
Image curtesy – (Nayak and Hazra, 2011; Parab and Bhalerao, 2010)
Common Pitfalls to avoid while applying statistical tests on clinical trial data(Charan and Saxena, 2012)
- Managing Missing Data – Missing data on large number of trial patients leads to skewing of results on either side. Care must be taken while dealing with a data set which has large number of missing entries. The integrity of clinical trial rest on completeness of the data.
- Applying wrong statistical test – Every statistical test has its own requirements, when the data fulfils the requirement of the test then only the results from the tests can be trustworthy. In cases, wherein data doesn’t satisfy the condition of normality non-parametric tests should be test of choice.
- Reporting of test results – A test result for continuous variable must be reported with Mean or Median, P Value, and 95% Confidence interval. While the test results for categorical variable must be reported as Percentage with Sample Size
- Ad hoc Reporting of Subgroup Analysis – CONSORT clearly states that the trial protocol should clearly mention number of subgroup analysis to be performed and post hoc subgroup analysis should be avoided
- Charan, J., Saxena, D., 2012. Suggested Statistical Reporting Guidelines for Clinical Trials Data. Indian J. Psychol. Med. 34, 25. doi:10.4103/0253-7176.96152
- Dembe, A.E., Partridge, J.S., Geist, L.C., 2011. Statistical software applications used in health services research: analysis of published studies in the U.S. BMC Health Serv. Res. 11, 252. doi:10.1186/1472-6963-11-252
- Gupta, S.K., 2011. Intention-to-treat concept: A review. Perspect. Clin. Res. 2, 109–112. doi:10.4103/2229-3485.83221
- Moher, D., Hopewell, S., Schulz, K.F., Montori, V., Gøtzsche, P.C., Devereaux, P.J., Elbourne, D., Egger, M., Altman, D.G., 2010. CONSORT 2010 Explanation and Elaboration: updated guidelines for reporting parallel group randomised trials. J. Clin. Epidemiol. 63, e1–e37. doi:10.1016/j.jclinepi.2010.03.004
- Nayak, B.K., Hazra, A., 2011. How to choose the right statistical test? Indian J. Ophthalmol. 59, 85–86. doi:10.4103/0301-4738.77005
- Parab, S., Bhalerao, S., 2010. Choosing statistical test. Int. J. Ayurveda Res. 1, 187–191. doi:10.4103/0974-7788.72494
- Parikh, T.B., Nanavati, R.N., Patankar, R.V., Pn, S.R., Bisure, K., Udani, R.H., Mehta, P., 2007. Fluconazole Prophylaxis against Fungal Colonization and Invasive Fungal Infection in Very Low Birth Weight Infants. Indian Pediatr. 44, 830–837. doi:10.1.1.554.3205
- Pearce, N., 2012. Classification of epidemiological study designs. Int. J. Epidemiol. 41, 393–397. doi:10.1093/ije/dys049
- Schulz, K.F., 2010. CONSORT 2010 Statement: Updated Guidelines for Reporting Parallel Group Randomized Trials. Ann. Intern. Med. 152, 726. doi:10.7326/0003-4819-152-11-201006010-00232
- Wassertheil-Smoller, S., Kim, M.Y., 2010. Statistical Analysis of Clinical Trials. Semin. Nucl. Med. 40, 357–363. doi:10.1053/j.semnuclmed.2010.04.001