We carried out a cross-sectional internet survey targeting physicians who self-accessed the survey homepage that was advertised through mailing lists, public medical journals. The survey did not involve a sampling process using means such as membership lists of specific medical organizations. For the survey, cooperation was sought from a non-specific pool of physicians. Further, it was anonymous in nature, and accessing the survey homepage was regarded as consent for survey participation. Encouragement toward accessing the homepage was limited to advertising through physician mailing lists, academic journals, and commercial medical journals. Four mailing lists were used for the advertisement: "Total Family Care" mailing list comprising approximately 2,500 primarily independent practitioners and primary care physicians, "Internist" mailing list comprising approximately 1,000 of the board members of the Japanese Society of Internal Medicine, "pEBM" mailing list comprising primarily evidence-based medicine (EBM) physicians, and "EML" mailing list comprising primarily emergency care providers. Journal advertisements were printed in bimonthly and biweekly Japanese medical journals in general medical fields with an emphasis on those for internal medicine.
The survey questions investigated the awareness regarding the withholding or withdrawal of potentially "life-extending treatment" in three case scenarios pertaining to medical intervention, namely, Case 1, Case 2, and Case 3. These three scenarios concerned judgment for the initiation/withholding of tube feeding for an elderly individual in a stroke-induced comatose state with a high potential for long-term prolongation, judgment for the attachment/withholding of a respirator in a patient with an identical status to the above patient with the additional occurrence of severe pneumonia, and judgment for the discontinuation/withdrawal of artificial feeding when a patient is in a prolonged comatose state for more than 6 months and the withdrawal of tube feeding has been requested by the patient's family (Appendix). Based on these three scenarios, the survey sought responses as to whether the available treatment options should or should not be withheld or withdrawn. The survey also used an analogous method to seek responses concerning actual actions in routine practice. Further, the survey also assessed whether physicians viewed two particular medical interventions as either "life-sustaining treatment" or not these two interventions were the continuation of artificial feeding through a gastrostomy for a patient in the third aforementioned scenario and the attachment of an artificial respirator when this patient developed severe pneumonia and would likely require more than 7 days until separation from the respirator could be undertaken.
In addition to the case scenario questions, we inquired the extent to which physicians make use of resources such as conferencing, consultation with clinical ethicists, and application to ethics committees when faced with difficult cases pertaining to ethical judgment.
Data input was carried out through an Internet homepage created specifically for the survey research, and electronic mail was not used. Physicians were asked to input their age, sex, number of years since graduation, and specialty; however, physicians did not provide any other personally identifying information. To achieve complete anonymity of personal information in the research, the server storing the response data was set up in a data center unaffiliated with the researchers. Information obtained by the researchers from the data center was completely anonymized, and researchers were entirely unable to obtain the internet protocol (IP) address of the respondents or other such information. For furthering the efforts to prevent the identification of individuals, the survey was carried out completely on a volunteer basis with no acknowledgements or incentives provided. The survey was opened on January 10, 2005 and remained open until March 31, 2005.
After all the survey mechanisms were complete, the researchers analyzed the anonymous data. To cleanse the data of the possibility of the same physician responding multiple times, data with identical answers for physician age, gender, field of practice, and employing institution and having a 75% or greater concordance in responses to the other questions were treated as responses from the same physician. In these cases, only data from the initial access were selected, and data from the second and subsequent accesses were deleted. In addition to descriptive statistics for each question, the statistical analysis included the calculation of kappa values for concordance between awareness and actual practice of withholding or withdrawal of specific treatments in each scenario and for concordance in responses across scenarios. The discrepancies in judgment-related awareness of the treatments were also compared by physician characteristics. The relationships between the attitudes with regard to the judgments in cases 1, 2, and 3 and the physician characteristics and experiences concerning ethical matters were analyzed using a logistic regression model. Odds ratios (OR) and their 95% confidence intervals (CI) were calculated.
The conduct of the research was approved by Tokyo Medical Center Ethics Committee in November, 2004.
Answers to all your Biology Questions
