The Evolution of Randomised Control Trials


It is estimated that one third of therapies are useless.  The difficulty is determining which third.  It is one of the major legacy issues, from a previous era, facing modern medicine.  Poor quality medical treatment is costly but more importantly it is potentially harmful.  It is now widely accepted that the best way to deliver best care to the patient into the future is through the practice of evidence-based medicine 1,2.  The upstream activity that provides the best evidence is the RCT.

The RCT has grown in stature decade by decade since first described by Bradford Hill in the 1940s.  The first randomised curative trial 1948, tested the efficacy of Streptomycin for curing pulmonary tuberculosis3. After 6 months therapy, there was significantly reduced morbidity and mortality in the Streptomycin group.  The design and methodology of this study helped to lay the foundation for future RCTs.   

The advent of multiple pharmaceutical inventions in the 1950s, antibiotics, anti-hypertensives, antipsychotics, required new measures to test their efficacy and safety.  The discovery of the teratogenic effects of Thalidomide, described in 1961, was a further catalyst to the introduction of rigorous evaluation of the safety of new therapies.  Thalidomide, an agent that was being prescribed for nausea and vomiting in pregnancy, was first marketed in Germany in 1957.  It caused devastating limb abnormalities in the fetus.  It is estimated that 10,000 children were affected, and that 2,000 died.  The US, however, remained unscathed.  Francis Oldham Kelsey, a reviewer for the FDA, refused on a number of occasions to grant a license for Thalidomide.  She had concerns regarding its safety, despite that it had been approved in Canada and 20 European countries.  She was influenced by her previous research which had indicated that some anti-malarial medications were capable of crossing the placenta.  Her reservations about the drug were subsequently shown to be well founded.  Her action saved countless thousands of US children from death and disability.  She was subsequently awarded the President’s Award for Distinguished Federal Civilian Service by John F. Kennedy.  In the years after the Thalidomide controversy, detailed evaluation of all new medications was commenced.  In the current era, the process is very searching.  Only 10% of all drugs entered into human trials are subsequently approved. 

By 1970 the FDA had mandated that new pharmacological agents underwent RCTs prior to approval.  The industry embraced the new changes and became an increasingly important sponsor of RCTs.  Their participation in RCTs surpassed Government sponsored and research funding bodies.  RCTs range greatly in magnitude.  They may be confined to one centre or involve multiple sites across many countries.  Many require a large staff, complex organization, and major funding.  At the extreme end of the spectrum, a phase 3 pharmaceutical RCT may cost $30 million.

A number of issues and reservations relating to RCTs have been addressed or need to be addressed.  An early concern was that of reporting bias.  Reporting bias occurs when the direction of the results of research influences its dissemination.  Dickersin and Chalmers4 have stated that it is unethical not to publish the results of an RCT.  Participants in clinical research are assured that their involvement will contribute to scientific knowledge.  In 2004 the international group of editors stated that clinical trial registration should be mandatory.

There is the tendency of editors and journals to accept RCTs with positive results and reject those with negative outcomes.  The odds of a positive paper being accepted is 2 or 3 times greater.  This practice defeats one of the main objective of RCTs, the eradication of ineffective therapies.  Changing this practice is challenging.  In comparing the proportion of negative RCTs published between the time periods 2002/3 and 2007/8 remained relatively unchanged 29% vs 34% across 3 major journals (JAMA, N Engl J Med, Lancet)5. The failure of researchers to submit their results is the major issue rather than the rejection by journals.

The RCT principle was seriously challenged during the 1980s AIDS epidemic.  Critically ill patients suffering from AIDs were deeply concerned that the RCT process would delay their acquisition of the new therapies.  The roles of the doctor and the scientist became conflicted.  The lobby group won special concessions including the use of surrogate end points.  Some critics remained concerned that this relaxation of the rules would lead to a deregulation that could be exploited by the drug industry.

Bothwell1 et al point out that on occasions the RCT fails to change clinical practice.  The reasons for the failure include the skepticism of physicians.  Another is the robust defense of the pharmaceutical industry in relation to its product.

RCTs in surgery present their own unique challenges.  It is suggested that the RCT has a more limited role in assessing surgery treatments than for drug interventions.  There are particular methodological challenges6.  The placebo-sham methodology has been used in some studies.  One good example was a placebo controlled evaluation of arthroscopic surgery for osteoarthritis of the knee, which showed no benefit compared with the sham procedure7

Sham controls cannot be used for major surgery which limit the feasibility of blinded studies.  Other factors include the skill level and experience of the operators.  Only 3.4% of articles in leading surgical journals are RCTs.


JFA Murphy




  1. Bothwell LE, Greene JA, Podolsky SH, Jones DS. Assessing the goal standard-lessons from the history of RCTs.  N Engl J Med 2016;374:2175-2181
  2. Woodcock J, Ware JH, Miller PW, McMurray JV, Harrington DP, Drazen JM. Clinical trials series. N Engl J Med 2016;374:2167
  3. Marshall G, Blacklock JWS, Cameron C, Bradford Hill A, et al. Streptomycin treatment of pulmonary tuberculosis. BMJ 1948;Oct:769-782
  4. Dickersin K, Chalmers I. Recognising, investigating and dealing with incomplete and biased reporting of clinical research: From Francis Bacon to the WHO. J R Soc Med 2011;104:532-538
  5. Brooks GA, Bowles GW, Jamieson DB et al. Reporting of negative randomized trials in 3 medical journals.  Int J Clin Med. 2013;4:108-113.
  6. Cook JA. The challenges faced in the design conduct and analysis of surgical randomised controlled trials. Trials 2009
  7. Moseley JB, O’Malley K, Petersen NJ, et al. A controlled trial of arthroscopic surgery for osteoarthritis of the knee. N Engl J Med 2002;347:81-88