Biomedical subjects
Publications and source records attributed to D Oakes.
Many mathematical representations are possible both for the frequency distribution of survival time and for the effect on that distribution of explanatory variables. A short review is given of the main types of model, the techniques available for model choice, the consequences of assuming a particular form and methods for assessing goodness of fit. Effects of model misspecification are discussed briefly.
A study was conducted to compare the "productivity" of a cohort of research grant applicants selected by peer review to be scholars of The Leukemia Society of America (now The Leukemia & Lymphoma Society) with a matched cohort of applicants not so selected during the period 1981 to 1990. One hundred and twenty-four scholars and 124 nonfunded applicants were studied. Two bibliometric variables and their derivatives were examined from the Institute of Scientific Information database: the number of papers published and the number of citations to those papers. Published papers were measured through December 31, 1999, and citation counts to these papers through December 31, 2000. Scholars published 10,301 papers through the period of observation and nonfunded applicants published 6442 papers. Scholars' papers were cited 536,283 [corrected] times, whereas nonfunded applicants' papers were cited 245,586 times. The mean citations per paper were 52 for scholars and 38 for nonfunded applicants. The papers published per scholar, citations per scholar, and citations per paper per scholar were significantly greater than the corresponding measures for nonfunded applicants (P < 0.0001 in each case). Scholar's papers were cited 30% more often, whereas nonfunded applicants were cited 10% more frequently, than a comparison group of scientists publishing in the same journal in the same year. High-impact papers, e.g., papers that were cited more than 200 times, were nearly three times as frequent among scholars (494 papers) as among nonfunded applicants (173 papers). This difference was highly significant. The good (better than baseline) performance of nonfunded applicants may be a reflection of self-selection among the applicant pool for this competitive award; the more productive performance of the scholars is probably the result of the selection decisions made during the peer-review process.
BACKGROUND AND OBJECTIVES: Among the various treatment modalities for condyloma acuminatum, excisional cold-blade surgery appears excellent but it has been little studied and little used, particularly for lesions not located in the perianal area. GOALS: To examine the efficacy and complications of scissors excision of single anogenital warts. STUDY DESIGN: Retrospective analysis of single warts completely excised with scissors for the purpose of biopsy before patient entry in a randomized, placebo-controlled study of the efficacy and safety of various parenteral interferons in combination with cryotherapy. RESULTS: Of 152 patients entered in the main study, 85 patients were available for analysis. At 4 and 16 weeks after excision, 16 of 85 (19%) and 14 of 68 (21%) of the excised lesions recurred. After at 6 least months of follow-up, 2 of 11 (18%) of the excision sites demonstrated some evidence of pigmentation changes. CONCLUSIONS: Scissors excision of single anogenital warts has a high rate of success and acceptable long-term side-effects.
Our objective was to study the placebo response in Parkinson's disease (PD). We conducted a literature search in which placebo response was measured in all studies of PD from 1969 to April 1996. Strict criteria were defined to include or exclude published reports in our survey. The Parkinson Study Group database for Deprenyl & Tocopherol Antioxidative Therapy of Parkinsonism (DATATOP) was reviewed and placebo "responders" were compared to placebo "nonresponders" by age, race, religion, level of education, duration of PD, and gender. A significant difference between the efficacy of placebo and that of active drug was reported in 61% (22 of 36) of the articles meeting the required criteria; DATATOP analysis showed no statistically significant epidemiologic differences between 140 placebo responders and 58 placebo nonresponders except in PD effect on current job. Although there is clearly a placebo response in PD patients, our review suggests that the variation in placebo response does not correlate with demographic factors such as age, gender, religion, level of education, or duration of PD.