[Show abstract][Hide abstract] ABSTRACT: Background/Aims: Research and surveillance work addressing ectopic pregnancy (EP) rely largely on ICD diagnostic codes and CPT procedure codes available from automated data sources. However, cases identified in this way may not be true EP cases, as these codes may also be used to rule out, follow up or note a history of EP. Through the use of additional automated data on treatment, procedures, and other aspects of care, we developed a classification algorithm that could improve the accuracy of EP case identification. Methods: Using Group Health automated data files, we initially identified 2,632 potential EP episodes occurring in women aged 15-44 years during 1988-2007 using ICD and CPT codes for EP and for surgical or laparoscopic treatment of EP. Chart reviews to verify EP status were conducted on 130 potential EP cases in the algorithm development dataset and on 150 in the algorithm validation dataset. Using additional information on demographic factors, other diagnosis and procedure codes, treatment modalities, site of care (ambulatory vs. inpatient), and laboratory data available with each EP episode, we conducted a classification and regression tree (CART) analysis to create a case finding algorithm for EP. Results: From the CART analysis, the case-finding algorithm for EP contained three main predictors: at least two encounter dates with an EP diagnosis or procedure code during an episode; treatment with methotrexate; and presence of an ICD -9 code of 633.1, 633.10 or 633.11 for tubal pregnancy. Sensitivity for the development and validation sets, respectively, was 95% and 91% while specificity was 78% and 83%. The EP misclassification rate using the new algorithm was 10.7% in the development set and 11.5% in the validation set compared to 32% when EP cases were originally defined using EP diagnosis and procedure codes. Conclusions: The CART-derived algorithm for identifying EP cases was highly sensitive, had good specificity, and misclassification rates were notably improved over the original case identification techniques. Additional pharmacy and encounter data available in many health plan databases can markedly improve the accuracy of EP identification. In particular, the identified predictors in the algorithm are available in the CRN VDW, and it would be of interest to test this algorithm at other HMORN sites.
[Show abstract][Hide abstract] ABSTRACT: To evaluate the feasibility and efficacy of population-based outreach strategies to improve genital Chlamydia trachomatis (CT) screening in men.
In a randomized trial, male enrollees ages 21-25 (n = 8820) were selected from the automated files of Group Health Cooperative and randomized to: a letter + test-request card for a CT urine home sampling kit (arm 1, n = 2940); a letter + mail-back sampling kit (arm 2, n = 2940); or a usual care control (arm 3, n = 2940). One reminder was sent to arms 1 and 2. The outcome was CT testing rates in the 4 months postrandomization.
105 of 2940 (3.6%) men in arm 1 and 230 of 2940 (7.8%) in arm 2 returned mailed specimens. All 335 respondents were sexually experienced, 43% had >2 sex partners in the past year, and 80% reported no genitourinary symptoms. Compared to arm 3, the relative risk of being tested was 5.6 (95% confidence interval (CI) 3.6-8.7) for arm 1 and 11.1 (95% CI 7.3-16.9) for arm 2. Arm 2 was significantly more likely to be tested than arm 1. CT prevalence for mailed-back specimens was 1.0% (1 of 105) for arm 1 and 2.6% (6 of 230) for arm 2; 70% of all positive intervention tests were from mailed samples.
Both strategies resulted in significantly higher CT testing than usual care, but the intervention response rate was low (5.7% overall). Direct kit mailing performed best. In US populations, the value of mailed outreach strategies to men must be considered in the context of other CT screening priorities.
Sex Transm Dis 12/2007; 34(11):837-9. · 2.59 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: To examine whether the presence of diabetes alters the risk of acute urinary tract infection (UTI) in postmenopausal women.
A case-control study of the Group Health Cooperative of Puget Sound (GHC), a staff-model nonprofit health maintenance organization in Washington State, was conducted. Subjects were women aged 55-75 years who had been members of GHC for at least 1 year and who had had an acute symptomatic UTI within the preceding month. Laboratory files were used to identify women with a urine culture that grew > or =10(5) colonies of a urinary pathogen. Medical records were reviewed to confirm the presence of acute, clinically symptomatic UTI. Control subjects were randomly selected from the GHC enrollment file, screened to remove women with recent UTI, and frequency matched to cases by age within 2 years. An interviewer ascertained self-reported clinician-diagnosed diabetes. Diagnosis of diabetes was confirmed by the GHC diabetes registry. A subsample of women underwent measurement of postvoid residual bladder volume (n = 748) and culture of vaginal flora (n = 454).
Of the 901 case and 913 control subjects, diabetes was reported in 13.1 and 6.8%, respectively. The health plan diabetes registry confirmed the diagnosis in 92% of women who self-reported the condition. The age-adjusted odds ratio (OR) for UTI in relation to self-reported clinician-diagnosed diabetes was 2.2 (95% CI 1.6-3.0). Adjustment for frequency of sexual intercourse and history of UTI had little effect on this estimate. Compared with nondiabetic women, higher UTI odds were seen in subjects who used oral hypoglycemic agents (OR 2.9 [95% CI 1.7-5.1]) and insulin (2.6 [1.5-4.6]) but not in subjects with untreated diabetes or diabetes treated by lifestyle changes (1.3 [0.7-2.3]). No significant difference was seen in the OR for UTI in diabetic women with disease of shorter duration (<10 years, OR 1.9) or longer duration (> or =10 years, OR 2.6) or in relation to HbA(1c) level. Similar microbiologic pathogens were seen in diabetic and nondiabetic women. No significant differences were seen by diabetes status in mean postvoid residual bladder volume or vaginal flora.
Diabetes under pharmacologic treatment is associated with increased risk of clinically apparent UTI in postmenopausal women.
Diabetes Care 11/2002; 25(10):1778-83. · 7.74 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Staphylococcus saprophyticus is the second most common cause of urinary tract infection (UTI) in young women. Relatively little is known about risk factors for this infection including exposure to vaginal spermicides, which increases the risk of UTI caused by Escherichia coli.
We conducted a case-control study in a large health maintenance organization Case patients were sexually active young women with acute UTIs caused by S saprophyticus identified from computerized laboratory files during 1990 to 1993. Population-based control patients were randomly selected from the organization's enrollment files. Exposures such as sexual activity and contraceptive practice were determined by interview.
Of 1299 eligible women, 66% (96 case patients and 629 control patients) were interviewed. Case patients were more often unmarried and were more sexually active. Ninety-nine percent of case patients and 57% of control patients reported previous UTIs. Exposure to any type of condom during the previous year was reported by 53% of case patients and 31% of control patients. Exposure to spermicide-coated condoms during the previous month was associated with a higher risk of UTI (odds ratio [OR], 3.8; 95% confidence interval, 1.4-10.3). The OR for exposure during the previous year ranged from 2.2 (95% confidence interval, 1.0-4.8) for less than once weekly to 6.05 (95% confidence interval, 2.2-16.6) for more than twice weekly. In multivariate analyses, younger age (OR, 0.97 per year), intercourse frequency (OR, 1.2 per weekly episode), prior UTI (OR, 3.3), and frequency of exposure to spermicide-coated condoms (OR, 8.4 for more than once weekly and 10.9 for more than twice weekly) were independent predictors of UTI. Among women exposed to spermicide-coated condoms, 74% of UTIs caused by S saprophyticus were attributable to this exposure.
Spermicide-coated condoms were associated with an increase risk of UTI caused by S saprophyticus. Because sexual activity and spermicide exposure are important risk factors for UTI caused by both S saprophyticus and E coli, it is likely that they share a similar pathogenesis.
Archives of Internal Medicine 03/1998; 158(3):281-7. · 11.46 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Diaphragm/spermicide use increases the risk of urinary tract infection (UTI). To determine whether spermicide-coated condoms are also associated with an increased risk of UTI, the authors conducted a case-control study at a large health maintenance organization in Seattle, Washington. Cases were sexually active young women with acute UTI caused by Escherichia coli, identified from computerized laboratory files during 1990-1993. Age-matched controls were randomly selected from the enrollment files of the plan. Of 1,904 eligible women, 604 cases and 629 controls (65%) were interviewed. During the previous year, 40% of the cases and 31% of the controls had been exposed to any type of condom. The unadjusted odds ratio for UTI increased with frequency of condom exposure from 0.91 (95% confidence interval (CI) 0.65-1.28) for weekly or less during the previous month to 2.11 (95% CI 1.37-3.26) for more than once weekly. Exposure to spermicide-coated condoms conferred a higher risk of UTI, with odds ratios ranging from 1.09 (95% CI 0.58-2.05) for use weekly or less to 3.05 (95% CI 1.47-6.35) for use more than once weekly. In multivariate analyses, intercourse frequency (odds ratio (OR) = 1.14 per weekly episode), history of UTI (OR = 2.64), and frequency of spermicide-coated condom exposure (OR = 3.34 for more than once weekly and 5.65 for use more than twice weekly) were independent predictors of UTI. Spermicide-coated condoms were responsible for 42% of the UTIs among women who were exposed to these products.
American Journal of Epidemiology 10/1996; 144(5):512-20. · 4.78 Impact Factor