The era of accountability.
To a certain extent, the era of accountability represents a fresh beginning. Payers, in particular, have felt that from this point forward there should be limited diffusion of a technology before adequate clinical trials have been performed. The gold standard of clinical trials has always been the randomized trial. How many journal articles conclude with the phrase, "preliminary results appear promising and this therapy warrants further investigation in randomized clinical trials," yet the randomized trial is never forthcoming? There is the hope that, if only we had the data, we would know the answers, and medical practice would proceed on firmer ground.
However, the experience with one technology, home uterine activity monitoring (HUAM), has shown that this hope is naive. Never has a technology been so thoroughly studied. Since its introduction in the late 1980s, there have been some eight randomized controlled clinical trials, meta-analyses, and policy statements by various groups, such as the American College of Obstetrics and Gynecology (ACOG). Even with this overwhelming wealth of data, this technology remains controversial.
HUAM is a diagnostic technology that is designed to provide earlier detection of preterm labor in women at high risk. These women can then be treated promptly with the hope that progression to preterm birth can be averted and likewise the immense costs associated with prolonged stays in neonatal intensive care units. The technology consists of a portable device that a woman can use at home to record contractions. On a daily basis, the woman can transmit these recordings over the telephone to a nurse, who then interprets the recordings and can give reassurance and advice about whether to see a physician. On the surface of it, the design of a randomized trial seems simple. The device is non-invasive and diagnostic, so unlike surgical or other therapeutic procedures, there are no major ethical issues regarding randomization, and the therapy could easily be blinded by either dummying a device or simply not using the transmitted recordings as part of treatment decisions.
However, it quickly became apparent that the trial design was extremely complex, not in terms of the randomization scheme, but in the patient selection criteria and the study outcomes chosen. To some degree, the clinical use of any diagnostic device is held captive to events on either side of it, i.e., the patient selection criteria, and then, after the diagnostic device is used, how the patient is treated and the success of that treatment. HUAM would only have an impact on those women who actually did develop preterm labor, which is difficult to predict. About 50 percent of women experience preterm labor with no identifiable risk factor, and even of those who appear to be at high risk, i.e., a previous history of a preterm birth, only a percentage will develop preterm labor again. Therefore, from the beginning, it was recognized that HUAM would be a waste of time and money in many women, depending on the patient selection criteria. Should these patients be included in the statistical analysis or discarded? Their inclusion would make it more difficult to detect a statistically significant decrease in the preterm births. To overcome this, studies with very large numbers of patients would be required.
Even in those women who did develop preterm labor, some believe that HUAM can be judged to be a total success only if the treatment is successful, i.e., if tokolysis results in a decrease in preterm birth. However, some studies have used intermediate endpoints, such as the degree of cervical dilation at the time of diagnosis. Advocates of HUAM have argued that it is inappropriate to tie a therapeutic outcome to a diagnostic technology, citing the analogy that it would be unreasonable to require a thermometer to cure an infection, rather than simply measure the fever. In response, HUAM critics would say that it is inappropriate to measure a fever with an expensive device (i.e., HUAM) if there is no evidence that the treatment works, or if a cheaper device is available.
Finally, what component of the HUAM system was responsible for an effect, if any--was it the daily nursing contact, or was it the objective evidence provided by the recorded uterine contractions? Some hypothesized that there might be a benefit from the assurance and advice given by the nurses, and that the recorded contractions added little to any potential effect.
The subsequent flurry of randomized trials have fanned out from these three basic points of controversy. There have been trials that have analyzed only those women who developed preterm labor, trials that looked at the rate of preterm birth, some that looked at the intermediate outcome of cervical dilation, others that looked at the total HUAM package (both the recording device and the nurse contact). Finally, there have been those that attempted to isolate the individual contribution of either the device or the nurse. Most of the trials showed a positive outcome in some parameter. Trial sizes varied, from less than 100 to, most recently, a trial that included more than 1,000 patients. There is a device that has been FDA approved on the basis of a trial that isolated the contribution of the crevice alone, analyzed only those who developed preterm labor, and looked at the intermediate outcome of cervical,dilation. Critics of HUAM pointed out that selecting the last two parameters severely biased the results toward a beneficial effect of HUAM.
All of these trials have been completed within the space of eight years, a remarkable accomplishment. However, as each trial was published, a lively debate ensued in the literature as to the validity of the trial design and whether there was adequate data to support the widespread adoption of this technology. At this point, some five organizations have published policy statements and recommendations that universally conclude that HUAM has not been shown to be an effective method of reducing the rate of preterm birth in high-risk mothers. These organizations include: American College of Obstetrics and Gynecology, the US. Health Preventive Service Task Force, the Agency for Health Care Policy and Research, the National Institute of Child Health and Human Development, and the American Medical Association, through its technology assessment program. Nevertheless, HUAM has been used by many obstetricians.
Certainly, other medical technologies have been embraced wholeheartedly on the basis of much thinner evidence but, in this case, there seems to be resistance on the part of the policy-making OB/GYPI community. Some have suggested that OB/GYNs have been chastened by the experience with routine fetal monitoring at the time of birth. This was a technology that was widely accepted for many years in the absence of clinical trials, which ultimately showed that routine fetal monitoring was largely unnecessary. Another suggestion has been that OB/GYNs are wary of establishing a premature standard of care that could raise liability issues. In other words, if a preterm infant was born to a mother who did not use the HUAM device, could the physician be held liable? However, there is also the strong possibility that the OB/GYN community, at least those in policy-making positions, have embraced the era of accountability and are holding HUAM to a new, higher standard.
Elizabeth Brown, MD, is Director of Technology Assessment and Clinical Guidelines at Aetna Health Plans, Aetna Life and Casualty, in Chicago, Illinois. The opinions expressed in this column are those of the author and not necessarily those of Aetna Health Plans. Dr Brown can be reached at 312/441-3292.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||clinical trials and outcomes data|
|Date:||Aug 1, 1996|
|Previous Article:||Interactive media enhances medicine.|
|Next Article:||The time is now.|