Search for a command to run...
The Internet is an accessible medium for information transfer, which makes the World Wide Web (www) component of the internet a very powerful vehicle for providing educational materials to the public. Although most of the so-called medical websites are not standardized, patients and physicians are increasingly using the internet as a source of medical information (Burneo, 2006). This is a reflection of the overall availability of computers and access to the www. The Internet is a resource that can rapidly reach a wide geographic population. This would include patients and physicians that otherwise may not have access to clinical trials information. The idea of using internet for clinical research is not new. But, its use in research has been basically limited to matching services mainly for oncology and HIV trials (Ohno-Machado et al., 1993; Wei et al., 2004; Metz et al., 2005). These have been shown to be effective capturing those patients interested in a specific trial, and from the investigator's perspective, as they have seen prescreened patients based on the inclusion and exclusion criteria of the trial. This bypasses some of the major roadblocks that have been identified in the problem of clinical trials recruitment (Metz et al., 2005). But from my point of view, that would be it. Bergin et al. also suggest randomization but without blinding, something of major concern. This idea would not work as randomization does not always succeed in its goal of achieving groups with similar prognosis. Neurologists in charge of the process may make mistakes that compromise randomization or they may encounter bad luck. Randomization does not provide the guarantees that the two groups (treatment and control groups) will remain prognostically balanced. Differences in patient care other than the intervention under study can also bias the results. Clinicians gain greatest confidence in study results when investigators document that all co-interventions that may plausibly impact on the outcome are administered more or less equally in treatment and control groups. The absence of such documentation is a much less serious problem if the neurologists are blind to whether patients are receiving active treatment or are part of the control group. Blinding eliminates the possibility of either conscious or unconscious differential administration of effective interventions to treatment and control groups. Furthermore, the lack of concealment would be another major problem with what Bergin et al. suggest. If those making the decision about patient eligibility are aware of the arm of the study to which the patient will be allocated, they may systematically enroll sicker (patients with prolonged seizures and/or patients with high frequency of seizures) or less sick patients to either treatment or control groups. This behavior will defeat the purpose of randomization and the study will yield a biased result (Schulz et al., 1995). The idea remains a great one, but only I believe applicable to the initial process of recruitment for clinical trials or for follow up purposes.