Using winter 2009-2010 to assess the accuracy of methods which estimate influenza-related morbidity and mortality.

We used the winter of 2009-2010, which had minimal influenza circulation due to the earlier 2009 influenza A(H1N1) pandemic, to test the accuracy of ecological trend methods used to estimate influenza-related deaths and hospitalizations. We aggregated weekly counts of person-time, all-cause deaths, and hospitalizations for pneumonia/influenza and respiratory/circulatory conditions from seven healthcare systems. We predicted the incidence of the outcomes during the winter of 2009-2010 using three different methods: a cyclic (Serfling) regression model, a cyclic regression model with viral circulation data (virological regression), and an autoregressive, integrated moving average model with viral circulation data (ARIMAX). We compared predicted non-influenza incidence with actual winter incidence. All three models generally displayed high accuracy, with prediction errors for death ranging from -5% to -2%. For hospitalizations, errors ranged from -10% to -2% for pneumonia/influenza and from -3% to 0% for respiratory/circulatory. The Serfling and virological models consistently outperformed the ARIMAX model. The three methods tested could predict incidence of non-influenza deaths and hospitalizations during a winter with negligible influenza circulation. However, meaningful mis-estimation of the burden of influenza can still result with outcomes for which the contribution of influenza is low, such as all-cause mortality.

Association between obesity and vulnerability and serologic response to influenza vaccination in older adults.

BACKGROUND: Serologic response to influenza vaccination declines with age. Few other host factors are known to be associated with serologic response. Our objective was to determine whether obesity and vulnerability independently predicted serologic response to influenza vaccination. METHODS: Adults ≥ 50 years were recruited during the 2008-2009 influenza season. Subjects provided pre- and post-vaccination sera for measuring antibody titers to 2008-2009 vaccine components. Body mass index (BMI) was calculated as weight (kg)/height (m(2)). Data were collected on vulnerability using the vulnerable elders survey (VES13). Logistic regression evaluated the associations between obesity and vulnerability and the serologic response to vaccination (both seroprotection and seroconversion), adjusting for gender, age, comorbidities, pre-vaccination titer, and site. RESULTS: Mean (± standard deviation) age of 415 study subjects was 65 ± 10 years; 40% were obese. Mean BMI was 29 ± 5.6 kg/m(2); mean VES13 was 1.6 ± 1.8. The proportions of subjects who seroconverted and had seroprotective titers were 40% and 49%, respectively, for A/Brisbane/59 (H1N1); 73% and 80% for A/Brisbane/10 (H3N2); and 34% and 94% for B/Florida. Modified VES-13 (score 0-10, with 10 being most vulnerable) was not associated with seroprotection against H1N1 or H3N2, and VES-13 was directly associated with seroconversion to H1N1 but not H3N2 or B. Obesity (BMI ≥ 30 kg/m(2) vs. BMI 18.5-30 kg/m(2)) was not associated with seroprotection for H1N1 or H3N2; obesity was directly associated with seroconversion to H3N2 but not H1N1 or B. Age was inversely associated with seroprotection and seroconversion against H1N1 and with seroconversion to influenza B. CONCLUSION: Based on this sample of older healthy subjects, there were no consistent relationships between VES 13 or obesity and either seroprotection or seroconversion to three influenza vaccine antigens.

NIH consensus development statement on management of hepatitis B.

OBJECTIVE: To provide health care providers, patients, and the general public with a responsible assessment of currently available data on the management of hepatitis B. PARTICIPANTS: A non-DHHS, nonadvocate 12-member panel representing the fields of hepatology and liver transplantation, gastroenterology, public health and epidemiology, infectious diseases, pathology, oncology, family practice, internal medicine, and a public representative. In addition, 22 experts from pertinent fields presented data to the panel and conference audience. EVIDENCE: Presentations by experts and a systematic review of the literature prepared by the Minnesota Evidence-based Practice Center, through the Agency for Healthcare Research and Quality. Scientific evidence was given precedence over anecdotal experience. CONFERENCE PROCESS: The panel drafted its statement based on scientific evidence presented in open forum and on published scientific literature. The draft statement was presented on the final day of the conference and circulated to the audience for comment. The panel released a revised statement later that day at http://consensus.nih.gov. This statement is an independent report of the panel and is not a policy statement of the NIH or the Federal Government. CONCLUSIONS: The most important predictors of cirrhosis or hepatocellular carcinoma in persons who have chronic HBV are persistently elevated HBV DNA and ALT levels in blood. Other risk factors include HBV genotype C infection, male sex, older age, family history of hepatocellular carcinoma, and co-infection with HCV or HIV. The major goals of anti-HBV therapy are to prevent the development of progressive disease, specifically cirrhosis and liver failure, as well as hepatocellular carcinoma development and subsequent death. To date, no RCTs of anti-HBV therapies have demonstrated a beneficial impact on overall mortality, liver-specific mortality, or development of hepatocellular carcinoma. Most published reports of hepatitis therapy use changes in short-term virologic, biochemical, and histologic parameters to infer likelihood of long-term benefit. Approved therapies are associated with improvements in intermediate biomarkers, including HBV DNA, HBeAg loss or seroconversion, decreases in ALT levels, and improvement in liver histology (Table). Although various monitoring practices have been recommended, no clear evidence exists for an optimal approach. The most important research needs include representative prospective cohort studies to define the natural history of the disease and large RCTs of monotherapy and combined therapies, including placebo-controlled trials, that measure the effects on clinical health outcomes. Table. Criteria Useful in Determining for Whom Therapy is Indicated: Patients for whom therapy is indicated: Patients who have acute liver failure, cirrhosis and clinical complications, cirrhosis or advanced fibrosis and HBV DNA in serum, or reactivation of chronic HBV after chemotherapy or immunosuppression; Infants born to women who are HBsAg-positive (immunoglobulin and vaccination). Patients for whom therapy may be indicated: Patients in the immune-active phase who do not have advanced fibrosis or cirrhosis. Patients for whom immediate therapy is not routinely indicated: Patients with chronic hepatitis B in the immune-tolerant phase (with high levels of serum HBV DNA but normal serum ALT levels or little activity on liver biopsy); Patients in the inactive carrier or low replicative phase (with low levels of or no detectable HBV DNA in serum and normal serum ALT levels); Patients who have latent HBV infection (HBV DNA without HBsAg). We recommend routine screening for hepatitis B of newly arrived immigrants to the United States from countries where the HBV prevalence rate is greater than 2%. Screening will facilitate the provision of medical and public health services for infected patients and their families and provide public health data on the burden of disease in immigrant populations. The screening test should not be used to prohibit immigration.

Population-based incidence of human granulocytic ehrlichiosis in northwestern Wisconsin, 1997-1999.

The incidence of human granulocytic ehrlichiosis (HGE) in the upper Midwest is uncertain. Active surveillance for suspected HGE was conducted from 1997 through 1999 in a 13-county region of northwestern Wisconsin. Suspected HGE cases were classified, according to the national case definition, as confirmed, probable, or not HGE. In total, 112 confirmed cases and 30 probable cases of HGE were identified. The median age of the 142 case patients was 56 years, and 92 (65%) were male; 111 (78%) were residents of the surveillance region. The mean annual incidence of confirmed and probable HGE was 9.3 cases per 100,000 residents; there was no increase from 1997 to 1999. The incidence was highest among persons > or =50 years old and residents of Washburn County. The incidence of HGE in this region exceeded prior estimates, but it was lower than the reported incidence in areas of endemicity in Connecticut.

A community intervention trial to promote judicious antibiotic use and reduce penicillin-resistant Streptococcus pneumoniae carriage in children.

OBJECTIVE: Inappropriate use of antibiotics is common in primary care, and effective interventions are needed to promote judicious antibiotic use and reduce antibiotic resistance. The objective of this study was to assess the impact of parent and clinician education on pediatric antibiotic prescribing and carriage of penicillin-nonsusceptible Streptococcus pneumoniae in child care facilities. METHODS: A nonrandomized, controlled, community intervention trial was conducted in northern Wisconsin Clinicians. Clinic staff received educational materials and small-group presentations; materials were distributed to parents through clinics, child care facilities, and community organizations. Prescribing data were analyzed for 151 clinicians who provided primary pediatric care; nasopharyngeal carriage of penicillin-nonsusceptible S pneumoniae was assessed for 664 children in the baseline period (January-June 1997) and for 472 children in the postintervention period (January-June 1998). RESULTS: The median number of solid antibiotic prescriptions per clinician declined 19% in the intervention region and 8% in the control region. The median number of liquid antibiotic prescriptions per clinician declined 11% in the intervention region, compared with an increase of 12% in the control region. Retail antibiotic sales declined in the intervention region but not in the control region. Among participating children in child care facilities, there were no significant differences in antibiotic use or penicillin-nonsusceptible S pneumoniae colonization between the intervention and control regions. CONCLUSIONS: A multifaceted educational program for clinicians and parents led to community-wide reductions in antibiotic prescribing, but in child care facilities, there was no apparent impact on judicious antibiotic use or colonization with drug-resistant S pneumoniae. Longer follow-up time or greater reductions in antibiotic use may be required to identify changes in the pneumococcal susceptibility.

A randomized trial of zinc nasal spray for the treatment of upper respiratory illness in adults.

PURPOSE: We performed a randomized, double-blind, placebo-controlled trial to evaluate the efficacy of 0.12% zinc sulfate nasal spray for reducing the duration and severity of acute upper respiratory infections. SUBJECTS AND METHODS: Patients with acute onset of upper respiratory illness of less than 24 hours' duration were eligible for the study. A nasopharyngeal swab was obtained at the time of enrollment for viral culture. Participants were randomly assigned to receive either 0.12% zinc sulfate or isotonic placebo spray. The medication was administered as two inhalations in each nostril four times a day. Each patient completed a diary card twice a day to record oral temperature, symptoms, and adverse effects. Symptoms were scored as absent (0), mild (1), moderate (2), or severe (3). RESULTS: One hundred eighty-five subjects volunteered to participate, and 160 met the criteria for enrollment. The median duration of all symptoms was 7 days in both groups (P = 0.45), and the median duration of nasal symptoms was 6 days in both groups (P= 0.12). After adjustment for baseline differences in severity, patients receiving zinc had a significant reduction in the total symptom score (P= 0.02) and the nasal symptom score (P= 0.02) on day 1, but not on any of the other days. Adverse effects were mild and had no significant association with the use of zinc. A respiratory virus was identified in 9 of the 160 participants; 6 of these were rhinovirus. CONCLUSION: A low concentration of zinc sulfate nasal spray had no effect on the duration of the common cold.

Tickborne infections as a cause of nonspecific febrile illness in Wisconsin.

Lyme disease, human granulocytic ehrlichiosis (HGE), and babesiosis are tickborne infections that are indigenous to Wisconsin. To assess their importance as a cause of nonspecific fever, we recruited patients with febrile illness at 10 clinics in northwestern Wisconsin from May through August of both 1997 and 1998. Eligible patients had a temperature >38.0 degrees C but no rash or other localizing source. Acute and convalescent serological tests were performed for Borrelia burgdorferi, Babesia microti, and Ehrlichia equi; polymerase chain reaction was performed to detect granulocytic Ehrlichia rDNA. Seventeen (27%) of 62 eligible patients had laboratory evidence of tickborne infection, including 7 (11%) with probable Lyme disease only, 8 (13%) with HGE only, and 2 (3%) with apparent coinfection. No patients with Babesia infection were identified. Patients with and without tickborne infection were similar with regard to age, sex, symptoms, history of tick bite, and outdoor exposure. The results suggest that tickborne infections are an important cause of nonspecific febrile illness during the tick season in northwestern Wisconsin.

Ixodes scapularis (Acari: Ixodidae): abundance and rate of infection with Borrelia burgdorferi in four state parks in Wisconsin.

Four state parks located in Lyme disease endemic regions of Wisconsin were surveyed for the presence of Ixodes scapularis Say during May and June of 1998 by drag sampling along hiking trails. Nymphal abundance varied between parks, with the average number of nymphs encountered in 1 h ranging from 6.2 +/- 3.8-47.1 +/- 36.3 (mean +/- SD). Questing nymphs were tested for the presence of Borrelia burgdorferi by culture in BSK medium and 7-12% was found to be infected. The average risk of encountering an infected nymph (entomologic risk index) ranged from 0.5 to 5.2 infected nymphs per hour. The highest entomological risk index was recorded from a small island park in northwestern Wisconsin during the last week in May (8.0 infected nymphs per hour). These results indicate a lower risk for human Lyme disease exposure in Wisconsin state parks in comparison with highly endemic areas of the northeastern United States.

The effect of a community intervention trial on parental knowledge and awareness of antibiotic resistance and appropriate antibiotic use in children.

BACKGROUND: Overuse of antibiotics for children's upper respiratory infections is widespread and contributes to the emergence of antibiotic-resistant bacteria. OBJECTIVE: To assess changes in knowledge and awareness regarding antibiotic resistance and appropriate antibiotic use after community-wide educational interventions to reduce inappropriate antibiotic use. DESIGN: Baseline survey conducted during June through July 1997 and postintervention survey of baseline participants during June through August 1998. SETTING: Communities in northern Wisconsin. PARTICIPANTS: Parents of 729 randomly selected children <4 years of age were called until 215 in each of the intervention and control areas were reached. Of the 430 baseline participants, 365 (85%) participated in the postintervention survey. INTERVENTION: Parent-oriented activities included distribution of materials and presentations. Physician-oriented activities included formal presentations and small group meetings. OUTCOME MEASURE: Change in awareness about antibiotic resistance and knowledge about antibiotic indications. RESULTS: A higher proportion of parents in the intervention area (53%) were exposed to 2 or more local educational messages, compared with the control area (23%). From the baseline to the postintervention survey, the percentage of parents with a high degree of antibiotic resistance awareness increased more in the intervention area (58% to 73%) than in the control area (60% to 65%). In the intervention area, there was also a larger increase in knowledge regarding appropriate indications for antibiotic use, compared with the control area. The proportion of parents who expected an antibiotic for their child and did not receive one declined in the intervention area (14% to 9%), while it increased in the control area (7% to 10%). In addition, the percentage of parents in the intervention area who brought their child to another physician because they did not receive an antibiotic decreased (5% to 2%), while it increased in the control area (2% to 4%). CONCLUSION: Parental knowledge and awareness about antibiotic indications and antibiotic resistance can be changed with educational interventions directed at parents and clinicians.

Antibiotic susceptibility of invasive Streptococcus pneumoniae in Wisconsin, 1999.

BACKGROUND: Streptococcus pneumoniae is a major cause of community acquired infections in the United States, and rates of antibiotic resistance have increased dramatically in the past decade. Statewide rates of pneumococcal resistance to penicillin and other antibiotics have not been previously reported in Wisconsin. To determine these rates, we assessed invasive pneumococcal isolates for reduced susceptibility to nine different antibiotics. METHODS: Pneumococcal isolates from blood, cerebrospinal fluid or other normally sterile body sites were submitted by 91% of laboratories that perform invasive bacterial cultures. Isolates were tested for susceptibility to penicillin, cefotaxime, ceftriaxone, levofloxacin, meropenem, erythromycin, vancomycin, sulfamethoxazole-trimethoprim and chloramphenicol. RESULTS: There were 409 invasive pneumococcal isolates identified in 1999 among Wisconsin residents, including 385 (94%) isolates from blood. The mean patient age was 42.5 years (range, < 1 year to 96 years), and 213 (52%) were male. Of the pneumococcal isolates, 24% were not susceptible to penicillin, including 10% with high level resistance. Isolates with reduced penicillin susceptibility were also likely to have reduced susceptibility to other antibiotics. Patients with penicillin nonsusceptible (intermediate and fully resistant) pneumococcal isolates were significantly younger (mean, 37.0 years) than those with susceptible isolates (mean, 44.3 years) (p = .04). The proportion of patients with a penicillin nonsusceptible isolate varied by region, ranging from 12.8% in northeastern Wisconsin to 35.5% in northern Wisconsin. CONCLUSIONS: The proportion of invasive pneumococcal isolates with penicillin resistance in Wisconsin is similar to other regions of the United States. Inappropriate antibiotic use contributes to the emergence of resistant pneumococcal infections, and educational efforts are underway to promote judicious antibiotic use in Wisconsin.

Clinical and epidemiological features of early Lyme disease and human granulocytic ehrlichiosis in Wisconsin.

To compare clinical features and assess risk factors for human granulocytic ehrlichiosis (HGE) and early Lyme disease, we enrolled patients in a case-control study during the 1996 and 1997 tick seasons. Clinical and demographic characteristics were assessed for patients with laboratory-confirmed cases of HGE or Lyme disease, and risk factors were compared with those of matched control subjects. We identified 83 persons with Lyme disease, 27 with HGE, and 11 with apparent coinfection. Unsuspected Ehrlichia infection was identified in 8 (13%) of 60 patients with Lyme disease. Patients with HGE were older and more likely to have fever, chills, or dyspnea than were those with Lyme disease only. Most patients with apparent coinfection did not have hematologic abnormalities. In the risk factor analysis, tickborne illness was independently associated with rural residence and camping. The clinical spectrum of HGE overlaps that of Lyme disease, and physicians in areas of endemicity should consider both diseases in treating patients with a compatible rash or febrile illness.

Physicians' practices and opinions regarding prenatal screening for human immunodeficiency virus and other sexually transmitted diseases.

BACKGROUND AND OBJECTIVES: Early prenatal diagnosis of sexually transmitted diseases (STDs), particularly human immunodeficiency virus (HIV), is critical for maternal and infant health. We conducted a survey to assess physicians' prenatal STD screening practices and opinions. STUDY DESIGN: A random sample of obstetricians and family physicians was selected from the Minnesota Medical Association directory to complete a standardized telephone survey. RESULTS: Eighty-three (86%) of 96 eligible obstetricians and 94 (95%) of 99 eligible family physicians completed the survey. Nearly all physicians recommend universal prenatal screening for syphilis (97%) and hepatitis B (99%); fewer physicians recommend prenatal screening for HIV (43%), chlamydia (26%), and gonorrhea (24%). Adjusting for physicians' specialty, female physicians were more likely than male physicians to recommend universal prenatal HIV screening (odds ratio [OR] = 2.1; 95% confidence interval [CI] = 1.1-4.2). Adjusting for physicians' ages, physicians with more than 20% uninsured/Medical Assistance patients were more likely than other physicians to recommend prenatal gonorrhea screening (OR = 3.1; 95% CI = 1.4-6.8); similar factors were associated with chlamydia screening. Although 89% of physicians supported universal prenatal HIV counseling and voluntary screening, the median percentage of prenatal patients screened for HIV was only 10%. CONCLUSIONS: Most physicians reported routinely screening prenatal patients for syphilis and hepatitis B. Although many physicians agreed with recommendations for universal prenatal HIV screening, their reported screening practices varied considerably from this approach.

Nonarteritic anterior ischemic optic neuropathy. A case-control study of potential risk factors.

OBJECTIVE: To determine the influence of certain potential risk factors on the occurrence of nonarteritic anterior ischemic optic neuropathy. DESIGN: Case-control using 2 independent control groups, one involving a geographically defined population and the other involving patients who underwent a routine comprehensive medical evaluation. SETTING: Multispecialty clinic in a rural setting providing primary, secondary, and tertiary care for residents of central and northern Wisconsin and the Upper Peninsula of Michigan. PATIENTS: Fifty-one patients older than 45 years with first ever acute nonarteritic anterior ischemic optic neuropathy. MAIN OUTCOME MEASURES: Potential risk factors defined using standardized definitions abstracted from the medical records, including diabetes, hypertension, hypercholesterolemia, coronary artery disease, tobacco use, chronic obstructive pulmonary disease, body mass index, hematocrit, and white blood cell count. METHODS: Conditional logistic regression analyses, first using a univariate analysis and then employing a multivariate analysis using a forward selection process. RESULTS: The geographically defined case-control multivariate analysis revealed that diabetes (odds ratio = 2.7, 95% confidence interval = 1.2-6.3, P = .02) and body mass index (odds ratio = 1.07, 95% confidence interval = 1.00-1.14, P = .08) were associated with case status. The comprehensive case-control multivariate analysis revealed that only diabetes (odds ratio = 5.0, 95% confidence interval = 1.4-17.3, P = .01) was a significant risk factor. The attributable risk estimation for diabetes was 0.21 and 0.27 for the geographically defined and comprehensive controls, respectively. CONCLUSIONS: Diabetes seems to be a major risk factor for the development of nonarteritic anterior ischemic optic neuropathy. The low attributable risk estimation suggests that factors other than diabetes are important in the development of nonarteritic anterior ischemic optic neuropathy or in predisposing individuals to it.

Prevalence of granulocytic Ehrlichia infection among white-tailed deer in Wisconsin.

Human granulocytic ehrlichiosis (HGE) is caused by an agent that is nearly indistinguishable from the veterinary pathogens Ehrlichia equi and Ehrlichia phagocytophila. The deer tick, Ixodes scapularis, is a vector of the HGE agent, and the white-tailed deer is the primary host for adult Ixodes ticks. We assessed the distribution of granulocytic Ehrlichia infection among deer living within (Wisconsin) and outside (western and southern Iowa) the geographic range of L. scapularis. Whole-blood samples were tested for HGE 16S ribosomal DNA (rDNA) by PCR, and E. equi antibody was detected by indirect immunofluorescence assay (IFA). Antibody titers of > or = 1:64 were defined as positive, and all positive samples were retested with a second lot of substrate antigen. E. equi antibody was present in 14 (8%) of 187 Wisconsin deer and 0 of 60 Iowa specimens (rate ratio undefined; P = 0.025). An additional 30 serum samples from Wisconsin deer were excluded because IFA results were discrepant between substrate lots. The reciprocal antibody titers ranged from 64 to 512 (geometric mean, 141) for positive samples. PCR results were positive for 27 (15%) of 181 Wisconsin deer. The prevalence of infection in northwestern Wisconsin deer was not significantly different from that in central Wisconsin deer, as determined by IFA and PCR. In two samples that were sequenced, the 16S rDNA was nearly identical to that of the granulocytic Ehrlichia species but distinct from that of Anaplasma marginale. The DNA sequences of the samples differed from the published sequences for E. equi, E. phagocytophila, and the HGE agent by 1 or 2 nucleotides (> or = 99.1% homology) at phylogenetically informative sites. Granulocytic Ehrlichia organisms in deer are widely distributed within the geographic range of L. scapularis in Wisconsin. Deer may serve as useful sentinels for areas where HGE transmission to humans may occur.

A population-based study of sexually transmitted disease incidence and risk factors in human immunodeficiency virus-infected people.

BACKGROUND AND OBJECTIVES: The Minnesota Department of Health conducts active surveillance for cases of human immunodeficiency virus (HIV) infection and passive surveillance for gonorrhea, Chlamydia trachomatis infection, and syphilis. The authors linked two computerized surveillance databases to assess gonorrhea incidence and risk factors for sexually transmitted disease (STD) acquisition among people with known HIV infection. STUDY DESIGN: People diagnosed with adolescent/adult HIV infection before 1993 and still alive as of December 31, 1994 were compared to people diagnosed with gonorrhea, chlamydial infection, or primary/secondary syphilis in 1993 or 1994. Records were matched on name, date of birth, and gender. The incidence of reported gonorrhea was calculated and risk factors for STD acquisition were examined. RESULTS: Thirty (1.3%) of 2,315 HIV-infected people were diagnosed with one or more STDs after HIV diagnosis (median interval: 3 years). There were 31 episodes of gonorrhea, seven episodes of chlamydial infection, and one episode of secondary syphilis. The gonorrhea incidence among HIV-infected people was high compared to the general population in Minnesota, even after stratifying by gender, age, and county of residence. STD acquisition was independently associated with female gender (odds ratio [OR] = 3.8; 95% confidence interval [CI] = 1.7, 8.3) and residence in Hennepin County (OR = 2.9; 95% CI = 1.2, 7.1), the most populous county in Minnesota. CONCLUSIONS: Linkage of STD and HIV surveillance data is useful as a sentinel for high-risk sexual behavior among HIV-infected people, and it can help identify individuals who require additional interventions to prevent HIV transmission. State and local health departments should consider linking these data sources to assess trends and allocate resources.

Factors associated with the geographic variation of reported Chlamydia infection in Minnesota.

BACKGROUND AND OBJECTIVES: In Minnesota, physicians have been required to report cases of Chlamydia trachomatis infection since 1985. The distribution of reported cases suggests that there is substantial geographic variation in the rate of chlamydia infection. GOAL OF THIS STUDY: We conducted chlamydia screening at selected sites and a survey of primary care physicians in counties with high and low rates of reported chlamydia infection. We hypothesized that chlamydia infections are uniformly distributed in nonmetropolitan areas, and the geographic differences in reported cases can be attributed to variable testing and reporting practices by physicians. STUDY DESIGN: The number of reported female chlamydia cases per 1,000 women was calculated for each rural Minnesota county in 1990. Fourteen counties with high and low rates of reporting were selected for further investigation. From September to December 1991, universal chlamydia screening was carried out at 11 clinics serving patients in these counties. A questionnaire was mailed to all primary care physicians in these counties. RESULTS: In 1990, the rate of reported chlamydia cases was 4.7/100,000 and 0.1/100,000 for women living in the high-reporting and low-reporting counties, respectively. Chlamydia infection was present in 5.5% and 9.7% of women screened at selected clinics in high- and low-reporting counties, respectively. Physicians in high- and low-reporting counties did not differ significantly in terms of age, gender, number of pelvic exams, or frequency of chlamydia testing. However, physicians in high-reporting counties were significantly more likely to test for chlamydia when evaluating mucopurulent cervicitis or salpingitis, and they were more likely to test a woman whose sex partner had urethritis. Physicians in high-reporting counties were also more likely to have reported a case of chlamydia.