Thursday, September 21, 2006
The Impact of Requiring Patient Authorization for Use of Data in Medical Records Research
The debate surrounding the appropriate and ethical use of individual medical records has crescendoed.[1-15] Like many medical dilemmas, this one requires a balance between individual rights and societal benefit. Individual rights advocates argue that medical record studies are not intended to benefit any one person, but they may put an individual at risk.[16-23] But even those people most concerned about the individual's right to privacy acknowledge the societal benefits that medical research has to offer.[16] To provide those benefits, the information used must be complete and unbiased. Complete access to existing data can help to ensure unbiased research results.[1]
In 1996, Minnesota translated this debate into a statute that went into effect on January 1, 1997 (Minnesota Statute 144.335: "Patient Consent to Release of Records," available by visiting the Journal's Web site at www.jfp.denver.co.us). This law requires each patient to sign a general authorization form (not informed consent) to release records for medical or scientific research.
The research community has speculated about the impact of the Minnesota statute and similar statutes on medical records research, but no data have been published.[24] Our study presents the rate of authorization refusal for a primary care multispecialty medical group and characterizes some of the differences between the people who refuse authorization and those who grant it.
METHOD
The Olmsted Medical Center (OMC) is a 75-clinician primary care multispecialty medical group in southeastern Minnesota, consisting of a large ambulatory care office in Rochester, Minnesota, and 12 branch offices in small communities throughout seven counties. The medical center includes a 65-bed acute care community hospital with an active obstetrical service, an outpatient surgery suite, an emergency department, and a full-time research department. For the past 35 years, the OMC has contributed data on all patient encounters to the Rochester Epidemiology Project, a population-based diagnostic index[25,26] housed in the Mayo Clinic and supported by funding from the National Institutes of Health.
SAMPLE
The study sample consisted of all new and established patients seen at the OMC for their first 1997 visit, during January or February, including scheduled office visits, emergency department visits, and admissions to the hospital and the outpatient surgery center. Since each ambulatory office keeps its own records, separate from those at the hospital, authorization is required at every site.
STUDY DESIGN
Each patient was asked by a department receptionist or hospital registration clerk to read and sign a general authorization form as a part of the normal registration procedure.(*) Patients who asked for additional information were given brochures. The authorization form was mailed to those patients who were unable to complete it during their emergency department visit or hospital stay. If a patient died before becoming stable enough to grant or refuse authorization, the next of kin was asked to sign the form. A parent or guardian was asked to sign the authorization form for children aged 16 years or younger and for adults legally unable to grant authorization.
The patient's authorization decision (granted, refused, or undecided) was entered into the electronic registration database of the clinic or hospital and was electronically linked to patient demographic data and the patient's stated reason for the first 1997 appointment. Patients who had a service or billing code but no corresponding authorization code were identified as not having been
asked about authorization.
ANALYSES
Simple descriptive statistics were used to estimate the proportion of patients granting authorization, refusing authorization, and remaining undecided (or not asked). The characteristics of patients refusing or granting authorization, and those either not asked or undecided, were assessed separately. Mantel-Haenszel tests were used to compare the proportions of patients granting authorization, refusing authorization, and undecided or not asked, across subgroups stratified by age, sex, and reason for appointment. Additional analyses were done to compare the two main groups, those patients granting and refusing authorization. Multivariate logistic regression analysis was used to describe the relationship between patients' characteristics and the decision to refuse authorization, after taking other characteristics into account.
Subscribe to Posts [Atom]