The survey instrument was developed after a comprehensive review of all published surveys prior to June 2001 and the completion of a pilot group of 10 patients from each identified site N= 80. After thorough review, collection and analysis of the pilot information, the final instrument was developed for administration. One of the more practical requirements of the survey was the ease of administration in the physician’s office, including time for completion. The survey’s median time for administration was 6.4 minutes (standard deviation +/- 1.8 minutes). The patient demographics including name, age, insurance type, ethnicity, educational status and number of prescriptions were obtained from the chart by the office staff prior to the administration of the survey. In the event of missing demographic information or patients at their first visit with the doctor, the information was obtained from the patient. Every patient was asked the number of medications that he/she was currently taking. The 12-item patient portion of the survey focused on exposure to DTC ads, role of the ads in their decisions to schedule an appointment or to ask the physician questions, previous experience with physicians and requests that were a result of the ads, and impact on their compliance (Table 1). The survey was confidentially administered to patients during their routine office visit by trained staff. In efforts to promote uniformity, the staff read the questions to the patients during their office encounters and collected the information. All patient visits were eligible for inclusion and only those patients that refused to answer the survey questions were excluded (n=2). Patients were administered the survey instrument only once for the duration of the study, irrespective of the number scheduled appointments during the study period.

Immediately following the patient visit, the physician completed a two-item questionnaire. The physician was blinded to the patient’s responses for the duration of the study (Table 2). The physician was never made aware of the patient’s responses even at the completion of this project. buy antibiotics canada

Table 1. PATIENT SURVEY INSTRUMENT

table1

The study time period was eight weeks of survey administration. However, during initial analysis of the surveys, the investigator’s noted some discrepancy between the patient’s intent to question the physician and the physician’s answers in the two-part questionnaire. Additionally, there were a small number of patients that did not indicate that they wanted to question the physician, but according to the physician’s answers, they were asked about a medication advertised. As a result, we expanded the original protocol to include a three-item follow up questionnaire specifically for this group (Table 3). This questionnaire was mailed directly to the patient with a self addressed stamped envelope for return within seven days requested. The research staff mailed the questionnaire at the time of data entry and notation of a discrepancy between the patient and the physician’s responses. Apcalis Oral Jelly

Table 2. PHYSICIAN EXIT QUESTIONS

table2

STATISTICAL ANALYSIS

Upon receipt of the completed surveys, we assigned an identifying number and entered the responses into a Microsoft Excel database. Only fully completed surveys were entered, there were no duplication of patient responders. The data was entered as the surveys were received from the sites. Categorical variables were analyzed by the chi-Square test for trend, and continuous data were analyzed by the unpaired Student’s Г-test to analyze significance between baseUne characteristics. A two-tailed P-value of less than .05 was considered statistically significant and only the data with P = < .05 are reported. Other standard statistical analyses were performed, including the Mann-Whitney test to test the difference of the percentage distributions in gender, ethnicity, education and number of prescription medications. All analyses were performed using the SAS statistical package for Windows, version 6.12 (Carey, NC). The data was evaluated within several subgroups such as age, gender, type and geographic region of practice, ethnicity, insurance status, and educational level of patients. The sites were analyzed individually and collectively to assess for any statistically significant difference between the respondents from the sites.

Table 3. FOLLOW UP (SENT VIA MAIL) TO PATIENTS THAT INDICATED THAT THEY WANTED TO DISCUSS AN AD WITH THE PHYSICIAN BUT DID NOT PER THE PHYSICIAN.

table3