Physicians' STD diagnosis and screening practices in the South.

Objectives: Sexually transmitted diseases (STDs) remain at high levels in the South compared with the rest of the nation. Physician diagnosis levels and screening behaviors fall among the elements about which more knowledge is needed to address these high levels. This article assesses Southern physicians' STD diagnosis histories and screening behaviors, focusing on curable STDs.

Methods: The sample included 1,306 physicians practicing in 13 Southern states and in the District of Columbia. These physicians formed part of a larger survey (n = 4,233) and answered questions concerning STD diagnosis history and screening behaviors. Analyses focus on chlamydial infection and gonorrhea individually, as well as composite statistics for gonorrhea, chlamydial infection, syphilis, pelvic inflammatory disease, trichomoniasis, and nongonococcal urethritis.

Results: Approximately 80% of physicians had diagnosed a curable STD, and 56% screened for any STD. The most common diagnosis techniques were culture and DNA probe. Several variables were individually associated with screening and diagnostic methods. Being female, African-American, or an obstetrician/gynecologist were associated with increased likelihood to screen for STDs in multivariate analyses.

Conclusions: Southern physicians were less likely to screen for STDs than their counterparts in other areas of the United States, although they were more likely to have diagnosed STDs. Results suggest that some targeted and evaluated screening practices may be useful in this area of the country.

Key Words: screening, sexually transmitted diseases, South

**********

Sexually transmitted diseases (STDs) remain a major public health problem in the United States. Of the top 10 most frequently reported diseases in 1995 in the United States, five were STDs, with approximately 15 million new cases of STDs in the United States occurring annually. (1) STDs present a serious health burden to society, including complications such as infertility, ectopic pregnancy, chronic pelvic pain, anogenital cancer, increased risk of human immunovirus (HIV) infection, and cardiovascular problems. (1) From 1973 through 1992, more than 150,000 women in the United States died as a result of complications associated with STDs, and 15% of all infertile American women are infertile because of pelvic inflammatory disease. STDs also present an economic burden on society, with an estimated cost of $17 billion annually. (2)

Curable STD rates in the United States are the highest in the developed world, with the highest rates within the United States occurring in the South (Figure). This area of the United States has consistently had the highest STD rates in the United States, and has had the highest rates of any region in the United States of gonorrhea and chlamydia for the past two decades to the present. (3) In 2001, chlamydial rates in the South ranged between 129.7 to 574.4 cases per 100,000 population, and gonorrhea rates ranged between 40.5 and 504, against a United States average of 278.3 and 128.5, respectively. (4) Six of the 10 states with the highest chlamydial rates were in the South, as were 7 of the 10 states with the highest gonorrhea rates. Finally, in 2001, all 16 Southern states had reported rates of primary and secondary syphilis that were greater than the Healthy People 2010 objective. (3)

The reasons for these elevated rates are comprised of many factors, among them, demographics, socio-economic factors, and access to health care. One part of the constellation of reasons lies with physician practices, and the purpose of this study is to determine the STD screening and diagnostic practices among Southern physicians. Importantly, this survey covers physicians in private as well as public settings. Brackbill et al (5) found that many STDs are evaluated, diagnosed, and treated in private primary care settings, as opposed to public health STD clinics. The current study, in addition to describing practices of Southern physicians in detail, compared them with physicians in the rest of the United States. Comparisons were principally aimed at screening rates and diagnostic practices.

[FIGURE OMITTED]

Materials and Methods

Sample

In 1999 to 2000, the Centers for Disease Control and Prevention and Battelle Centers conducted a national, probability-based survey of 7,300 physicians. Physicians were randomly selected from the five American Medical Association (AMA) specialties (general/family practice, obstetrics and gynecology, emergency medicine, pediatrics, internal medicine) that diagnose about 85% of STDs in the United States. Further eligibility criteria were that the physicians had seen patients between the ages of 13 and 60 years, and that they spent at least 50% of their time in direct patient care. Physicians received surveys via Federal Express, with repeat mailings at 4, 7, and 15 weeks to nonrespondents. We included $15 as an incentive for completing the survey. Overall, 70.2% of eligible physicians responded to the survey, results from which have been published elsewhere. (6-9)

The US Public Health Service divides the United States and its dependencies into four quadrants, the Northeast, West, Midwest, and South. Of the 4,233 respondents in the national sample, 1,306 (31.9%) came from the Southern quadrant (Figure) and formed the sample in the current analyses. At 20% of the sample, Texas physicians provided the most responses.

Analyses

Beyond descriptive statistics, tests were done to compare the associations between physician and practice characteristics, and diagnosis rates and screening practices. Bivariate associations were examined using a choice of: (1) Pearson correlations when both variables were continuous; (2) mean comparisons via t tests or analysis of variance (ANOVA) when one variable was continuous and one was categorical; and (3) chi-squares when both variables were categorical. Post hoc comparisons following the bivariate models were strictly exploratory and therefore a stringent criterion, the Scheffe test, was used. This test uses an alpha level adjusted for all pair-wise comparisons.

For multivariate models involving more than two variables, diagnosis rates (continuous) and screening (dichotomous) as the outcome variables were delineated. For diagnosis rates, analysis of covariance (ANCOVA) was used, which allows continuous and categorical predictors of diagnosis rates. For screening, logistic regression was used.

Results

Sample Description

A description of the personal, practice, and patient characteristics of physicians in the Southern quadrant can be found in Table 1. Compared with population statistics from the 2000 Census, Southern physicians were more likely to be white or Asian-American, and male. There were about half the number of African-American physicians as would be expected by chance, and a little over half the number of female and Latino/a physicians. Ages ranged from 25 to 90 years, and experience from 1 to 60 years, with the modal physician being about 45 years old with 17 years experience. These statistics are similar to physicians in the rest of the United States. (6) The exceptions were that Southern physicians spend an average of 45 hours per week in direct patient care, versus 41 hours for non-Southern physicians (P < 0.001) and more likely to be in public settings, 15.0% versus 11.7% (P < 0.01, both P values computed from Fisher exact test). Southern physicians also saw more patients per week on average, 106 versus 95 (t[4.018] = 4.27, P < 0.001). Reflecting the growth of cities in the South, a majority of physicians (53.8%) practiced in urban areas of 50,000 people or more, with only about 1 in 10 practicing in communities of 10,000 or fewer people. Primary care facilities and single specialty groups were the most common practice types.

Screening and Diagnostic Practices

Tables 2, 3, and 4 contain descriptions of sample physicians' diagnosis rates and screening practices. Of the sample, 91% of physicians stated they diagnosed gonorrhea and 90% chlamydial infections; however, 44% of physicians did not screen any patients for any STD. Roughly 3 in 10 screened females for either chlamydial infection or gonorrhea; roughly 1 in 10 screened males. Across the entire sample, about one in eight females (16.4%) fell into the approximate age range (13-25 yr) where screening of females for chlamydial infection is recommended. (10)

DNA probe and laboratory cultures were the most popular methods for diagnosing both gonorrhea and chlamydial infections, although physicians used a diversity of diagnostic measures. For chlamydial infections, 51% of responding physicians "always" used DNA probe, and 20% "always" used culture. For gonorrhea, the corresponding figures were 44% and 33%. Using a particular method to diagnose one STD was strongly correlated with using that method to diagnose the other (r = 0.68 for culture and 0.83 for DNA probe; both P < 0.001). After DNA probe and culture tests, no other method was used more than "sometimes," with urine-based polymerase chain reaction and ligase chain reaction (PCR/LCR) "usually" or "always" used by only 4.5% of physicians for diagnosing chlamydial infections, and 1.6% of physicians for diagnosing gonorrhea. Numbers of chlamydial infections and gonorrhea cases diagnosed were closely related (r = 0.77, P < 0.001) reflecting the epidemiologic similarities between these two STDs.

Correlations Between Respondent Characteristics and Screening and Diagnostic Practices

Bivariate associations involving screening practices. As indicated in Table 5, female physicians (66%) were more likely to screen patients for any STD than were male physicians (52%). Higher percentages of African-American physicians (76%) and Asian physicians (64%) screened patients for any STD than did white physicians (53%). The most substantial bivariate effect was for AMA specialty: a far higher percentage of obstetricians and gynecologists (95%) screened than any of the other physicians, which encompassed internists (41%), emergency room physicians (38%), general/family practitioners (63%), and pediatricians (30%). The size of the surrounding community was not associated with the likelihood of a physician screening for STDs, so it cannot be said that urban versus rural practice was a factor. Larger numbers of female patients, both overall and in the 13- to 25-year age range, were associated with greater likelihood of screening, but larger numbers of white patients and male patients between 13 and 25 years were associated with less screening. More physicians practicing in public settings screened patients than did physicians practicing in private settings, 71.1% versus 53.4%. Other factors captured in the study, but which were not associated with screening rates, were physician age, practice experience, percentage of black patients, and average patient load per week.

Bivariate associations involving STD diagnosis rates. Tables 3 and 4 show bivariate associations involving both diagnosis rates and diagnostic methods. As would be expected, the overall number of STDs diagnosed was correlated with average weekly patient load (r = 0.13, P < 0.001). The small size of this correlation (similar small correlations were found for chlamydial infections and gonorrhea individually) suggests that other factors also play a role in predicting diagnosis history. Two such factors were patient gender and race. However, physician characteristics, such as gender, race, and ethnicity were not associated with diagnosis rates.

Among other characteristics, primary specialty and practice location were both associated with overall numbers of STDs diagnosed. The clinic category was dropped from the latter analysis because it was composed of several different types of clinics, each with few respondents. The statistics based on all STDs applied to both gonorrhea and chlamydial infections individually, as shown in Table 3. According to Scheffe post hoc tests, the principal factor underlying these overall tests was the relatively high diagnosis rate by emergency medicine. Emergency medicine physicians diagnosed means of 62 to 135 more cases of STDs per year than all other specialties. Obstetricians and gynecologists diagnosed an average of 32 to 73 cases per year more than the remaining three specialties, none of which differed significantly from one another. This pattern was repeated, albeit with smaller mean differences for chlamydial infections and gonorrhea. This pattern was also reinforced by the Scheffe tests for practice site. Physicians practicing in hospital emergency rooms yielded the highest numbers of diagnoses, an average of 105 to 106 STD cases per year more than physicians in primary care facilities and ambulatory care facilities (including 23-25 more cases per year of gonorrhea and 22-24 more cases per year of chlamydial infection).

Practice type was only associated with the diagnosis of chlamydial infection (F[5, 1,194] = 2.78, and gonorrhea, F[5, 1,198] = 2.68, both P < 0.05) but not overall numbers of STDs diagnosed. For both STDs, Scheffe tests revealed the overall differences were principally due to lower numbers of diagnoses in solo practices compared with "other" settings, averages of 12 fewer cases of chlamydial infection per year and 8 fewer cases of gonorrhea. These differences are relatively small compared with the effects of practice site and primary specialty.

Composite models. Numerous bivariate associations were found, but some could be confounded with each other. Therefore, a number of multivariate models were conducted with diagnosis rates and screening behaviors as outcomes. Screening ("yes" versus "no") is a dichotomous variable, so logistic regression was used (Table 4). With screening for any STD as the outcome variable, the overall model was significant ([chi square][15] = 331.59, P < 0.001). Odds ratios for categorical variables remaining in the equation are presented in Table 4.

For diagnosis rates, the predictors are a mix of categorical (AMA specialty, practice location) and continuous (percent female and percent black patients, weekly patient load) variables, while the outcome is continuous. To incorporate this mix, ANCOVA was used with continuous predictors entered as covariates. The overall result was statistically significant (F[16, 981] = 19.26, R = 0.49, P < 0.001) with significant individual effects remaining for all variables entered except percentage of female patients. That is, the bivariate effects for AMA specialty and practice location were largely independent of one another (there was no statistical interaction). This means that emergency medicine physicians, regardless of whether they practiced in an emergency room or not, and physicians in emergency room locations, regardless of specialty, all diagnosed more STDs than other doctors. Identical results with gonorrhea and chlamydial diagnoses were found.

Comparisons to other physicians. Although this paper focuses on Southern physicians, we have noted their demographic similarity to other physicians. Comparisons of screening and diagnostic behaviors are also warranted. Although Southern physicians' average numbers of chlamydial (10.79) and gonorrhea (6.72) cases diagnosed per year did not differ statistically from other physicians' mean diagnoses (9.42 and 4.99, respectively), Southern doctors were more likely to have diagnosed at least one case of gonorrhea, 63.2% versus 50.2% (P < 0.001). Diagnosis rates for chlamydial infections were spread evenly by geographic area. Southern physicians were, however, slightly less likely to screen for STDs, 44.0% versus 39.9% ([chi square][1] = 6.00, P < 0.05). This pattern of lower screening rates was repeated for chlamydial infections and gonorrhea, but not syphilis. Regarding diagnostic methods, Southern physicians were less likely to use culture for gonorrhea (t[3,014] = 3.63, P < 0.001) and more likely to use DNA probe (t[3,068] = 2.90, P < 0.01). The same pattern held for chlamydial infection, with less use of culture (t[2,725] = 3.02, P < 0.01), and more use of DNA probe (t[3,067] = 3.28, P < 0.01).

Discussion

With some differences, such as somewhat higher weekly case loads, Southern physicians had largely similar characteristics to other physicians in the United States. Diagnosis of STDs was common among the sample, more common than screening asymptomatic patients (although more than half screened at least some of the time). Obstetricians/gynecologists and female physicians were most likely to screen patients, whereas emergency medicine physicians and those with high percentages of African-American patients diagnosed the most STDs. This discussion section concerns the implications of these findings, with differences other physicians noted where they add context.

Screening is the first question that deserves attention. In theory, the combination of high STD rates in the South and lower screening rates suggests an increase in screening is warranted. Barriers, however, include high case loads and lack of funding for universal screening. Therefore, a focus on more targeted screening may be more beneficial. Emergency medicine physicians diagnosed significantly more STDs than other disciplines, but they screened the least. Would routine STD screening by emergency medicine physicians yield enough cases to be worthwhile? Several studies have demonstrated that emergency departments may be an appropriate site for screening of gonorrhea and chlamydia, especially among 18- to 31-year-old patients. One study conducted in St. Louis, MO (11) found a 9.7% prevalence rate for gonorrhea and chlamydial infection among patients presenting to the emergency department, while another in Atlanta (12) found a prevalence rate of 9% for chlamydia and 6.5% for gonorrhea. Mehta et al, (13) in Baltimore, MD, found a prevalence rate of 13.6% for gonorrhea or chlamydia among 18- to 31-year-old patients in the emergency department. Mehta et al also collected cost data and found that screening in Baltimore emergency departments was also cost-effective. (14) In sum, emergency department screening, perhaps in the larger Southern cities, is a plausible means of cost-effective case-finding, and more so than in other areas of the country. One remaining caveat is that cost-effectiveness differs at different levels. A cost-effective strategy for a given state may not be so for an urban emergency department if the latter bears the costs of screening and the former the benefit of reduced transmission, fewer lost workdays, etc. This scenario suggests that some sort of reimbursement for emergency department physicians would be needed. Any tests of screening in emergency departments might profitably be extended to males, who are almost certainly (asymptomatic) reservoirs of infection for women, and who appear to be overlooked for screening in current practice (Table 3).

Another factor in the decision to screen, and in the prospective chances for success with a screening program in high morbidity settings, is the method used to screen. Southern physicians preferred to use DNA probe and culture for screening. DNA probe tests, which permit simultaneous testing for gonorrhea and chlamydial infections, were more popular among Southern physicians than elsewhere, and might facilitate a multi-STD screening program. Both this and culture, however, are considered invasive, and their use might inhibit patients who would otherwise agree to screening. Urine-based tests are much less invasive, but more expensive.

Pediatricians screened at a rate of 30% in this study. Although pediatricians traditionally are viewed as providing care for infants and children, 15- to 17-year-olds are included in pediatricians' scope of practice, and the STD epidemic may benefit from more screening among pediatricians, preferably based on taking an accurate and confidential sexual history from their older patients.

Specialty, race, and gender were associated with increased likelihood to screen for STDs in a multivariate model. Specifically, being female, African-American or Asian, or an obstetrician/gynecologist were most strongly associated, with the strongest association by far being medical specialty. This is consistent with the findings in a study by Lewis and Freeman (15) who found that women physicians screen for STDs more often, and with a previous study (7) published from this data set that found the highest rates of screening among obstetricians and gynecologists. Specialty is an understandable associate of screening because obstetricians and gynecologists are expected to screen at least their pregnant patients.

In contrast to prior studies, this study found the characteristics of the physician seem to function as a marker for screening, rather than characteristics of the patient. Specifically, the sex and race/ethnicity of the patient in this study did not matter when specialty and physician characteristics were taken into account (Table 4). Given existing racial and ethnic disparities in STD prevalence, it is more surprising to see physician characteristics associated with screening than it is to see patient race and ethnicity. Of course, it would be preferable if neither physician nor patient characteristics were associated with screening patterns. Instead, assessing behavioral risk factors for acquisition of STDs is an important part of history-taking, and often indicates when the need for screening exists. Although this paper does not address history-taking, other studies have evaluated how often physicians assess risk for STDs/HIV. (16) Results vary, with percentages between 10 and 40% of physicians routinely assessing risk factors for STDs/HIV. (15, 17-21) Were physicians or other staff to evaluate behavior more rigorously, the race, ethnicity, and other personal characteristics may lose their predictive value.

One potential limitation of the study is response bias. Those physicians that do screen for STDs or have an interest in STD prevention may have been more inclined to return the survey. Another limitation is recall bias. Some physicians may inaccurately describe their practices in relation to STD care, whether it be under-reporting or over-reporting.

Conclusion

Southern physicians have to be more sensitive to the need for screening, because STD rates in the South are higher than elsewhere in the United States and have been so for at least the past two decades. Screening rates are currently lower, while a larger proportion of Southern physicians have diagnosed STDs over the course of a year. Suggestions for improving STD health care, including case finding, center around carefully targeted and evaluated screening, including the use of behavioral assessment to guide screening decisions.

Table 1. Characteristics of Southern physicians in five specialties (a)

Respondent characteristics            No. (%)

Gender
  Male                                 934 (72.6)
  Female                               353 (27.3)
Race/ethnicity
  African-American/black                77 (5.9)
  American Indian/Alaska Native         13 (1.0)
  Asian                                121 (9.9)
  Native Hawaiian/Pacific Islander       1 (0.1)
  White                               1018 (77.9)
  Other                                 30 (2.3)
  Hispanic/Latino origin                98 (7.7)
Age (years)                           M = 45.76 SD = 10.27
Years in practice                     M = 17.34 SD = 10.40
Hours in direct patient care          M = 45.13 SD = 16.59
Primary specialty
  Emergency medicine                   117 (9.0)
  General or family practice           465 (35.6)
  Internal medicine                    267 (20.4)
  Obstetrics and gynecology            219 (16.8)
  Pediatrics                           238 (18.2)

Practice characteristics

Practice location
  Community of 1-10,000                140 (10.9)
  Town of 10,001-50,000                284 (22.9)
  City of 50,001-250,000               378 (29.4)
  City of 250,001 +                    314 (24.4)
Practice site
  Primary care facility                919 (70.7)
  Hospital emergency room              129 (9.9)
  Ambulatory care facility             123 (9.5)
  Clinic (community, public, urgent)    83 (6.4)
  Other                                 45 (3.4)
Practice type
  Solo practice                        332 (25.9)
  Single-specialty group practice      613 (47.8)
  Multi-specialty group practice       200 (15.6)
  Staff model HMO                       35 (2.7)
  Other managed care organization       32 (2.5)
  Other                                 70 (5.5)
Patient load per week                 M = 106.23 SD = 74.15
Patient gender (percent female)       M = 63.44 SD = 20.45
Patient race/ethnicity
  African-American/black              M = 26.09 SD = 20.71
  American Indian/Alaska Native       M = 0.64 SD = 3.44
  Asian                               M = 2.91 SD = 4.62
  Native Hawaiian/Pacific Islander    M = 0.28 SD = 2.94
  White                               M = 63.09 SD = 23.97
  Other                               M = 6.08 SD = 14.44
  Hispanic/Latino origin              M = 12.24 SD = 19.25

(a) M = mean; SD = standard deviation.
(b) Numbers do not always add up to 1,306 because not all physicians
answered all questions. Missing responses constitute < 2% of
responses to any given question.

Table 2. Percentage of Southern physicians screening asymptomatic
patients

                   Non-pregnant  Pregnant
           Males   females       females

Gonorrhea  10.26%  27.34%        29.40%
Chlamydia   9.26%  30.25%        29.17%

Table 3. STD diagnosis history (past year) (a)

       All bacterial
          STD         Chlamydia    Gonorrhea
Cases  No.  (%)       No.  (%)     No.  (%)

 0     113  (19.4)    305  (25.0)  451  (36.8)
 1-10  288  (45.8)    633  (54.3)  633  (51.7)
11-20  196  (12.9)    105   (8.6)   62   (7.1)
21-30  111   (5.6)     64   (5.2)   25   (2.0)
31-40   85   (5.0)     19   (1.6)   12   (1.0)
41 +   342  (11.4)     62   (5.1)   42   (3.4)

(a) Numbers do not always add up to 1.306 because not all physicians
answered all questions. Missing responses constitute < 2% of responses
to any given question. All bacterial STD includes gonorrhea, chlamydial
infection, syphilis, pelvic inflammatory disease, nongonococcal
urethritis, and trichomoniasis.

Table 4. Bivariate associations involving STD screening, diagnosis
rates, and methods

                                        Screening
                                        practices for
                                        any STD

Gender                                  [chi square](1) = 18.50 (a)
Race                                    [chi square](3) = 18.68 (a)
Private versus public practice setting  [chi square](1) = 20.16 (a)
AMA specialty                           [chi square](4) = 240.54 (a)
Weekly hours in direct patient care     t(1249) = 2.08 (b)
% female patients                       t(1245) = 13.14 (a)
% female patients between 13-25 yr      t(1232) = 6.75 (a)
% male patients between 13-25 yr        t(1227) = -3.89 (a)
% white patients                        t(1235) = -3.60 (a)
% Asian patients                        t(1236) = 3.68 (a)
% Hispanic patients                     t(1222) = 2.75 (c)

(a) P < 0.001
(b) P < 0.05
(c) P < 0.01

Table 5. Diagnosis rates (a)

                                           Chlamydial
                   All STD                 infection

AMA Specialty      F(4, 1130) = 44.16 (b)  F(4, 1216) = 30.15 (b)
% female patients  r(1126) = 0.14 (b)      r(1210) = 0.10 (b)
% black patients   r(1120) = 0.21 (b)      r(1201) = 0.17 (b)
Practice location  F(2, 1014) = 65.66 (b)  F(2, 1092) = 68.54 (b)

                   Gonorrhea

AMA Specialty      F(4, 1220) = 40.60 (b)
% female patients  r(1213) = 0
% black patients   r(1205) = 0.21 (b)
Practice location  F(2, 1096) = 107.96 (c)

(a) Missing responses constitute < 2% of responses to any given
question. "All STD" includes gonorrhea, chlamydial infection, syphilis,
pelvic inflammatory disease, nongonococcal urethritis, and
trichomoniasis.
(b) All values P < 0.001. except where noted.
(c) P < 0.01.

Table 6. Composite model of correlates of STD screening among Southern
physicians, screening for any STD (a)

Correlates (b)                       OR        95% CI

Physician gender  Male               Referent
                  Female             2.04      1.45-2.86
Physician race    White              Referent
                  Black              2.49      1.31-4.76
                  Asian              2.03      1.28-3.24
                  Other              1.47      0.78-2.75
AMA specialty     ER                 Referent
                  Family practice    3.14      1.96-5.02
                  Internal medicine  1.18      0.71-1.96
                  Ob/Gyn             20.83     7.87-55.56
                  Pediatrics         0.52      0.31-0.88
Practice setting  Private            Referent
                  Public             2.87      1.91-4.31

(a) OR, odds ratio: CI, confidence interval.
(b) n = 1,180 for this analysis.

Accepted January 8, 2004.

Copyright [c] 2004 by The Southern Medical Association

0038-4348/04/9707-0624

Please see Zakari Yusufu Aliyu's editorial on page 619 of this issue.

References

1. Institute of Medicine. The Hidden Epidemic: Confronting Sexually Transmitted Diseases. Committee on Prevention and Control of Sexually Transmitted Diseases. Washington, DC: National Academy Press; 1997.

2. American Social Health Association. Facts and answers about STDs. Available at www.ashastd.org. Accessed October 20, 2003.

3. Centers for Disease Control and Prevention. Tracking the hidden epidemics: trends in STDs in the United States 2000. Washington, DC: U. S. Government Printing Office, 2001.

4. Centers for Disease Control and Prevention. Sexually transmitted disease surveillance: 2001. Atlanta, GA: Department of Health and Human Services, 2002.

5. Brackbill R, Sternberg M, Fishbein M. Where do people go for treatment of sexually transmitted diseases? Fam Plann Perspect 1999;3:10-15.

6. St. Lawrence JS, Montano DE, Kasprzyk D, et al. STD screening, testing, case reporting, and clinical and partner notification practices: a survey of US physicians. Am J Pub Health 2002;92:1784-1788.

7. Hogben M, St. Lawrence JS, Kasprzyk D, et al. Sexually transmitted disease screening in the United States by obstetricians and gynecologists. Obstet Gynecol 2002;100:801-807.

8. McCree DH, Leichliter J, Hogben M, et al. Human Papillomavirus screening by U.S. physicians. Prev Med 2003;36:159-63.

9. McCree DH, Hogben M, St Lawrence JS. Differences among physician treatment practices for gonorrhea, syphilis, and chlamydia in the United States. Sex Transm Inf 2003;79:254-256.

10. U. S. Preventive Services Task Force. Guide to Clinical Preventive Services, 2nd Edition. Washington, DC: U. S. Department of Health and Human Services, 1996.

11. Todd CS, Haase CE, Stoner BP. Prevalence of asymptomatic gonococcal and chlamydial infections in an emergency department population. International Society for STD Research, Thirteenth Meeting, Abstract No. 47, July 11-14, 1999, Denver, CO.

12. McNagny SE, Parker RM, Zenilman JM, et al. Urinary leukocyte esterase test: a screening method for asymptomatic chlamydial and gonococcal infection in men. J Infect Dis 1992;165:573-576.

13. Mehta SD, Rothman RE, Kellen GD, et al. Unsuspected gonorrhea and chlamydia in patients of an urban adult emergency department: a critical population for STD control intervention. Sex Transm Dis 2001 Jan; 28(1):33-39.

14. Mehta SD, Bishai D, Howell MR, et al. Cost-effectiveness of five strategies for gonorrhea and chlamydia control among female and male emergency department patients. Sex Transm Dis 2002 Feb;29(2):83-91.

15. Lewis C, Freeman H. The sexual history-taking and counseling practices of primary care physicians. Western J Med 147;165-167.

16. Centers for Disease Control and Prevention. HIV Prevention Practices of Primary Care Physicians. MMWR Morb Mortal Wkly Rep 1994;42(51)988-992.

17. Lafferty WE, Downey L, Holan C, et al. Provision of sexual health services to adolescent enrollees in Medicaid managed care. Am J Public Health 2002 Nov;92(11):1779-1783.

18. Tao G, Irwin K, Kassler W. Missed opportunities to assess sexually transmitted diseases in US adults during routine medical checkups. Am J Prev Med 2000;18:109-114.

19. Gemson D, Colombotos J, Elinson J et al. Acquired immunodeficiency syndrome prevention: knowledge, attitudes, and practices of primary care physicians. Arch Intern Med 151;1102-1108.

20. McCance KL, Moser R, Smith K. A survey of physicians' knowledge and application of AIDS prevention capabilities. Am J Prev Med 1991;7(3):141-145.

21. Schuster MA, Bell RM, Peterson LP, et al. Communication between adolescents and physicians about sexual behavior and risk prevention. Arch Pediatr Adolesc Med 1996;150(9):906-913.

RELATED ARTICLE: Key Points

* Approximately 80% of Southern physicians had diagnosed a curable STD and 56% had screened for any STD.

* Being female, African-American, or an obstetrician/gynecologist were associated with increased likelihood to screen for STDs in multivariate analyses.

* The diagnosis pattern of results by specialty suggest that some targeted and evaluated screening practices may be useful in this area of the country.

Yolanda H. Wimberly, MD, MSC, and Matthew Hogben, PHD

From the Department of Pediatrics, Morehouse School of Medicine, and the Division Of STD Prevention, Centers for Disease Control and Prevention, Atlanta, GA.

Reprint requests to Yolanda Hill Wimberly, Morehouse School of Medicine, Department of Pediatrics, 720 Westview Drive SW, Atlanta, GA 30310-1495. E-mail: [e-mail]