With two months of research added to their already impressive résumés, the eight students participating in the Department of Family Medicine's Summer Student Research and Clinical Assistantship program (SSRCA) presented their findings to faculty, staff and friends in Alumni Hall, Madison, WI.

The SSRCA has been a fixture of the DFM for at least fifteen years and has been lead by Jon Temte, MD, PhD over the past decade. The program is designed to pair UW medical students and UW-DFM faculty who have mutual interests in research. The goals of the program are to:

• Provide medical students with an early exposure to Family Medicine
• Provide students with early exposure to clinical research and medicine in a variety of settings
• Stimulate and support research by Wisconsin family physicians and DFM faculty

The students apply for the assistantship positions and select from a listing of available faculty projects that they might find appealing. Through an interview process, faculty and students agree upon their matches for the summer.

As written by our summer researchers, the following is background information, methods used, results and conclusions of their research.

Respiratory Virus Surveillance in Wisconsin

Alexis Eastman and Jon Temte, MD, PhD used 14 years of viral epidemiology data, collected by the Wisconsin State Laboratory of Hygiene, to better characterize acute respiratory tract infections with known viral causes.

Background: Patient demographics and nasopharyngeal cultures were obtained, analyzed and recorded in a database. Since 1997, a checklist of associated symptoms has also been utilized. The system and symptoms have never been assessed.

Methods: In this study, statistical analysis utilizing Minitab software was performed on the data from 1997-2006 in order to examine two hypotheses:

• The first hypothesis was that the system performed accurately and to the specifications desired by the State Lab. (Mainly samples were collected during the autumn and winter seasons from patients with recent onset of symptoms, had an emphasis on children and were transported rapidly to the lab).
• The second hypothesis was that recorded symptoms would provide distinguishing and significantly different profiles for the viruses diagnosed. Basic statistical analysis was performed to investigate the first hypothesis, and ANOVA, and Two-way logistic regressions were performed on a list of 28 symptoms and factors to investigate the second.

Results: The first hypothesis appears to be supported by the data, but is still under investigation. The second hypothesis is supported by the evidence, with unique profiles being obtained for adenoviruses, influenza A, influenza B, rhinovirus, RSV, HSV, parainfluenza and enterovirus samples (p <0.001).

Conclusions: Symptoms were found to be both correlated and predictive of the assessed viruses, offering a tool for evidence-based clinical diagnosis of upper respiratory viruses during the autumn and winter seasons.

Awareness and Use of Nasal Irrigation Among Primary Care Physicians in the Wisconsin Research and Education Network: A Survey Study

Amy Bamber and David Rabago, MD hoped to determine physician awareness of Nasal Irrigation (NI) as a therapy for rhinosinusitis, the extent to which physicians use NI in their practice and physician satisfaction with NI. They developed a web-based survey to send to 180 Wisconsin physicians who are part of the WREN network this summer. The survey had been launched and initial results were not available at the time of presentation.

Mental Workload in Primary Care Practice: Contributions of Time Constraints and Problem Density

Betsy Doherty worked with Jon Temte, MD, PhD conducting a pilot study to better define the relationships between encounter problem density during primary care visits (number of problems presented per primary care visits), clinician mental workload, physician's estimates of medical error and quality indicators.

Background: The average number of problems per encounter in Family Medicine is 3.05 (Beasley, Ann Fam Med, 2004), and the average problem density is 10 -11 problems per hour (Temte, unpublished data). Patient complexity contributes to physician workloads, and problem density is one gauge of this complexity. In a systems-based patient care model, problem density may be one target of policy change to improve quality of care.

Methods: Two family physicians at UW-Wingra Clinic performed the NASA Task Load Index (TLX), a subjective multidimensional workload assessment, following consecutive patient visits during either one or three clinic shifts. A total of 33 patient visits were evaluated, with the additional data of problems per visit and duration of visit collected.

Results: The study revealed a mean TLX score of 36.72 (22.46-50.98, 95% CI) on a 100-point scale. Scores were similar between the two clinicians and among sessions, with the exception of one clinic session resulting in an atypically low TLX mean that corresponded to an unusually "light morning." Using regression analysis, significant correlations were found between number of problems and TLX score (0.66 P < 0.0001) and problem density and TLX score (0.736, P < 0.00001). The mean problem density by scheduled time was 10.93/hour (3.78-18.08m, 95% CI) and by actual time was 8.98/hour (4.14-13.82, 95% CI).

Conclusions: Increasing complexity with greater problem density may jeopardize the quality of patient care as well as quality of life of providers. This pilot study suggests that an instrument such as the NASA TLX may provide a reliable index of family physicians' mental workload. A future study that will concurrently examine markers of care quality and perceived error is underway.

Underage Drinking in Eau Claire County: Expanding Assessment & Capacity for the Reality Check 21 Program

Anna Graupner and Jennifer Eddy, MD of Eau Claire worked with state and county public health agencies, public health, hospitals and schools to assemble the most recent and accurate statistics regarding the costs, prevalence and effects of underage drinking in Eau Claire County. This initiative will provide local policymakers with accurate statistics as well as inform them about which approaches have been shown to be most successful in addressing this problem.

Background: Eau Claire County is located in northwest Wisconsin, which leads the nation in several alcohol-related statistics including rate of underage binge drinking. The EC City-County Health Department has worked via a comprehensive approach to address the issue and has formed multiple organizations, assessment and programs since 2001. A template was created in 2003 to characterize the "picture of alcohol use in ECC" with its 147 indicators.

Methods: The Needs and Resources Assessment Template was analyzed for missing data and available up-to-date data were recorded in a new template. 2005 PRIDE surveys and online data were used. The data indicators not found were categorized and amended with additional qualitative and quantitative questions. E-mails, phone calls and interviews were conducted with personnel from many different city and county government departments, as well as with health professionals from other organizations.

Results: Data were collected and/or updated for 116 out of 147 underage drinking indicators. Quantitative data were obtained from the EC Police Department, ECC Sheriff's Department, City Hall, Juvenile Detention and the EC Public Library. Qualitative data were gathered from EC Police Department, Sacred Heart Hospital Emergency Department, Juvenile Detention and ECC Traffic Clerk. Inquiries need to be made with the ECC Jail and Courts. Further research with the Police, Sheriff and emergency departments is desired.

Conclusions: Despite the diffuse nature of the research and the time limitations, a more complete picture of Eau Claire County's alcohol use and underage drinking indicators has been obtained. Many city and county departments, in addition to the public health department, collect data relevant to this issue. There is further potential for collaboration on public health initiatives relating to underage drinking with the hospitals' emergency departments and possibly the jail, juvenile detention and court programs.

Evaluations of the Diagnosis and Treatment of Pneumonia in Outpatient Primary Care Practices

Ann Regnery performed a retrospective chart review for 600 adult patients who have been diagnosed in an outpatient clinic with pneumonia, acute bronchitis or lower respiratory tract infection. She worked with Dennis Baumgardner, MD and Jennifer Evertsen, MS at the Milwaukee Campus. They evaluated both the criteria used by primary care physicians to empirically diagnose pneumonia and the criteria they use to order a chest radiograph to rule-out pneumonia.

Background: Many articles discuss the importance of ordering a chest x-ray to diagnose pneumonia. Chest x-rays are recommended when signs and symptoms or pneumonia are present, but it is unclear which signs and symptoms indicate pneumonia. The purpose of this study is to understand the clinical criteria used to diagnose and treat pneumonia compared to bronchitis and upper respiratory infections (URIs).

Methods: A retrospective chart review of 600 patients (18-80), 200 of each diagnosed with pneumonia, acute bronchitis or URIs within Aurora Health Care outpatient clinics. Chi-square and Kruskall-Wallis tests for univariate analysis and logistic regression for multivariate analysis were used.

Results: 608 patient charts were reviewed. Race, gender and COPD were significant predictors of pneumonia in both univariate and multivariate analysis. Dyspnea, chills, etc. were all significant individually. A stepwise regression identified rales or rhonchi, temperatures = 100, heart rate, rhinorrhea, cough and chest pain to best explain the variation in the diagnosis of pneumonia compared to bronchitis and URI (R2 = 44.11). Rales or rhonchi, dyspnea, rhinorrhea, chills, chest pain and temperatures = 100 were the best predictors (R2 = 38.22) of ordering a chest x-ray. The same variables remained significant as independent predictors of ordering a chest x-ray in binary logistic regression except chills and respiratory rate.

Conclusions: No clear prediction criteria are found to diagnose pneumonia compared to bronchitis and URI. A combination of signs and symptoms identified using the regression models above should be considered before ordering a chest x-ray or diagnosing pneumonia.

Bacterial and Fungal Growth in Breast Milk of Lactating Women with Breast and/or Nipple Pain

Sarah Allen joined Anne Eglash, MD, of the DFM Mount Horeb Clinic, to collect breast milk samples from lactating women and to develop reliable techniques to grow and identify bacteria and yeast from breast milk cultures. They compared rates of bacterial and yeast growth from breast milk among lactating women with and without chronic pain. They worked in conjunction with Richard A. Proctor, MD and Carol A. Spiegel, PhD of the Department of Medicine.

Background: Breast milk is well established as the ideal nutritional source for infants during the first year of life. There is a lack of information available to health professionals about breastfeeding difficulties including chronic breast and nipple pain. Women who experience these symptoms are left without information from their doctor and are more likely to wean prematurely. Previous studies show that breast milk cultures are often negative for bacterial growth; although when the women are treated with 6-8 weeks of antibiotics, their symptoms resolve. This study will aim to develop techniques to culture bacterial organisms from human breast milk.

Methods: Approximately 1-2 oz. of breast milk and one nipple culture will be collected from 40 lactating women, 20 with chronic breast pain who meet certain criteria and 20 who are asymptomatic. All milk samples will undergo culture techniques to optimize bacterial growth. In addition, women will be given a 14-21 question survey about their symptoms and pregnancy.

Resultswere pending at the time of presentation.

Care of International Adoptees by Family Physicians: A Clinical Practice Survey

Matt Hinderaker worked on a Nationwide Family Medicine International Adoption Clinical Practice Survey with Jon Temte, MD, PhD and third-year Madison resident Kristin Lerberg, MD. They surveyed the interest and comfort levels of family physicians throughout the United States in providing comprehensive health care to children, who were adopted internationally, and their families.

Background: Approximately 22,000 foreign children are adopted by U.S. families each year, but little is known about the attitudes and practices of physicians treating internationally adopted children.

Methods: We surveyed by mail 1,000 family physicians randomly selected from American Association of Family Physicians membership to determine their interest and comfort level in providing medical care for families with children who were adopted internationally.

Results: On average, family physicians deemed it "very appropriate" to provide comprehensive care for children who were adopted internationally; less than 2% thought these children should be referred to a specialty clinic. Physicians who practice more than 60 miles away from the nearest international adoption clinic tended to rate the appropriateness higher than physicians that practice in the same city as an international adoption clinic. Physician responses were statistically independent of the prevalence of international adoption in the physician's state of practice. Eighty-three percent of physicians responded that additional training would help them better care for internationally adopted children. The three most requested topics for additional training were:

• Medical exam, including laboratory and immunizations (55%)
• Pre-adoption counseling (53%)
• Review of international medical records (50%)

Women's Knowledge of Commonly Used Contraception

Sarah Hoffmann worked with Sarina Schrager, MD to conduct a survey of women of childbearing age who are waiting to visit with their DFM health care provider. They are interested in pursuing information about why, with the availability of contraceptive methods today, half of all pregnancies are still unintended. Sara Hoffman and Dr. Schrager asked the women about their knowledge of the various methods of available contraception and about their attitudes toward the most common contraceptive methods.

Background: Women in the U.S. can choose from a variety of safe and reliable methods of contraception. However, one half of pregnancies in this country are unintended, and half of unintended pregnancies end in abortion. There is a lack of research examining the knowledge women have toward various methods of contraception. The aim of this study is to learn more about what women today know about contraceptive methods, as this information will aid providers in addressing misconceptions and improving consistent use of contraception among their patients.

Methods: A questionnaire was given to English-speaking women between ages 18-40 in two outpatient family medicine clinics in Madison, Wisconsin assessing:

• Demographics
• Knowledge of risks and benefits of various contraceptives
• Information sources

Results: Approximately half of the 252 women studied were under the impression that condoms were 99% effective. Over half (58%) of women were unaware that birth control pills can decrease the risk of some types of cancer. Only 57% of women knew that condoms were not as effective as birth control pills. Over one quarter of women were using no method of birth control when given the questionnaire. The most common single source of information about birth control was the clinic, with 40% of women receiving information from two or more sources. There were no correlations between demographics, number of children, type of birth control method used and number of questions answered correctly, according to the Pearson Correlation Test.

Conclusions: Our findings demonstrate that there are serious misconceptions about the effectiveness of condoms among young women. A lack of knowledge about one important benefit of birth control (that it can decrease the risk of some types of cancer) is also apparent from the questionnaire data. Increased patient education by health care providers is necessary to emphasize the benefits and eliminate common misconceptions that patients may have about contraceptive methods.