A network-informed approach to investigating a tuberculosis outbreak: implications for enhancing contact investigations

BACKGROUND

To elucidate networks of Mycobacterium tuberculosis transmission, it may be appropriate to characterize the types of relationships among tuberculosis (TB) cases and their contacts (with and without latent TB infection) in addition to relying on traditional efforts to distinguish 'close' from 'casual' contacts.

SETTING

A TB outbreak in a US low incidence state.

OBJECTIVE

To evaluate whether social network analysis can provide insights into transmission settings that might otherwise go unrecognized by routine practices.

DESIGN

All adult outbreak-associated cases (n = 19) and a convenience sample of their contacts with and without latent TB infection (LTBI) (n = 26) were re-interviewed in 2001 using a structured questionnaire. Network analysis software was used to create diagrams illustrating important persons within the outbreak network, as well as types of activities TB cases engaged in with their contacts.

RESULTS

Drug use and drug sharing were more commonly reported among cases and their infected contacts than among contacts without LTBI. TB cases central to the outbreak network used crack cocaine, uncovering the need to focus control efforts on specific sites and persons involved in illicit drug use.

CONCLUSION

Outbreaks occur even in areas with low TB incidence, frequently among groups whose drug use or other illegal activities complicate control efforts. TB programs should consider the use of network analysis as a supplement to routine contact investigations to identify unrecognized patterns of M. tuberculosis transmission.

Contact tracing: comparing the approaches for sexually transmitted diseases and tuberculosis

SETTING

Literature review for the process of contact tracing for sexually transmitted diseases (STD) and for tuberculosis (TB), focusing on articles that report results of studies or commentary.

OBJECTIVE

To compare and contrast contact tracing in order to highlight emerging commonalities.

DESIGN

A descriptive review, based on Medline search with augmentation from other published and unpublished sources.

RESULTS

Contact tracing for STD and TB have some obvious differences because of differing routes of transmission, differing sensibilities required to work with the affected populations, a different potential for anonymous contacts, and a major difference in the epidemiologic value of biomarkers. Nonetheless, the convergence of these processes on disadvantaged populations where drug use and sexual activity are important social factors has engendered an increasing similarity.

CONCLUSION

A broadened approach to both, with greater attention to how ancillary contacts and associates may be of use in interrupting deeply embedded endemic disease transmission, deserves further study. Some newer approaches in the use of network-informed methods to elicit contacts and investigate the community dynamics of transmission may be of particular value in TB case investigation. These strategies will be enhanced by the availability of DNA fingerprinting, a powerful biomarker of recent Mycobacterium tuberculosis transmission and case association (a technology not available for STD contact tracing).

Risk network structure in the early epidemic phase of HIV transmission in Colorado Springs

This study describes the risk network structure of persons with HIV infection during its early epidemic phase in Colorado Springs, USA, using analysis of community-wide HIV/AIDS contact tracing records (sexual and injecting drug partners) from 1985 to 1999. Paired partner information from other STD/HIV programme records was used to augment network connections. Analyses were conducted with and without this supplemental information. The results suggest that a combined dendritic and cyclic structural network pattern is associated with low to moderate HIV propagation in Colorado Springs, and may account for the absence of intense propagation of the virus.

Sexual network structure as an indicator of epidemic phase

Ascertaining epidemic phase for a sexually transmitted disease (STD) has depended on secular trend data which often contain significant artefacts. The usefulness of sexual network structure as an indicator of STD epidemic phase is explored in an analysis of community wide genital chlamydia reports, with network analysis of interviewed cases and linked sexual partners, in Colorado Springs, USA, 1996 to 1999. In this period, the chlamydia case rate per 100,000 increased by 46%. Three quarters of cases (n=4953) were interviewed, nominating 7365 partners; these, combined with index cases, made up the 9114 persons in the network. Epidemiologic analysis of cases suggests that secular trend increases are artefactual. Network analysis supports this view: overall network structure is fragmented and dendritic, notably lacking the cyclic (closed loops) structures associated with network cohesion and thus with efficient STD transmission. Comparison of network structure with that of an intense STD outbreak (characterised by numerous cyclic structures) suggests low level or declining endemic rather than epidemic chlamydia transmission during the study interval. These observations accord with intuitive and stochastic predictions.

Birds of a feather: using a rotational box plot to assess ascertainment bias

BACKGROUND

Comparability of study participants with non-participants is customarily assessed by contrasting the distributions of sociodemographic characteristics. Such comparisons do not necessarily provide insight into whether or not participants of a given subgroup are similar to non-participants of the same subgroup. A geographical information system (GIS) may provide such insight by visually displaying the spatial distributions of participants and non-participants. In a previously reported study of heterosexuals at elevated risk for human immunodeficiency virus (HIV), traditional methods suggested distributional differences in the demographic characteristics of participants and non-participants.

METHODS

Based on residential address co-ordinates for each subgroup member, we used the subgroup's centroid as the origin and constructed a 360 degrees series of overlapping box plots of the distance of subgroups members to the origin, thereby producing closed polygons for each of the box plot demarcators.

RESULTS

These rotational box plots revealed similar geographical distributions for most participant and non-participant subgroups, with the exception of African-American men and women.

CONCLUSIONS

Observed differences resulted in part from the study design, and provided some insight into sampling problems encountered in social network studies. Based on Tobler's supposition that 'nearby things tend to be alike', the rotational box plot is a useful additional tool for investigating sample bias.

The Atlanta Urban Networks Study: a blueprint for endemic transmission

OBJECTIVE

To study prospectively social networks and behavior in a group of persons at risk for HIV because of their drug-using and sexual practices, with particular emphasis on the interaction of risks and concomitant network structure.

METHODS

A longitudinal study was conducted of 228 respondents in Atlanta, Georgia in six inner-city community chains of connected persons, interviewing primary respondents and a sample of their contacts every 6 months for 2 years. Ascertained were: HIV and immunologic status; demographic, medical, and behavioral factors; and the composition of the social, sexual, and drug-using networks.

RESULTS

The prevalence of HIV in this group was 13.3% and the incidence density was 1.8% per year. Substantial simultaneity of risk-taking was observed, with a high level of both non-injecting (crack, 82%) and injecting (heroin, cocaine or both, 16 30%) drug use, the exchange of sex or money for drugs by men (approximately 35%) and women (57-71%), and high frequency of same-sex sexual activity by men (9.4%) and women (33%). The intensity of interaction, as measured by network features such as microstructures and concurrency, was significantly greater than that observed in a low prevalence area with little endemic transmission.

CONCLUSION

The traditional hierarchical classification of risk for HIV may impede our understanding of transmission dynamics, which, in the setting of an inner-city population, is characterized by simultaneity of risk-taking, and moderately intense network interactions. The study provides further evidence for the relationship of network structure to transmission dynamics, but highlights the difficulties of using network information for prediction of individual seroconversion.

The effect of treating sexually transmitted diseases on the transmission of HIV in dually infected persons: a clinic-based estimate. Ad Hoc STD/HIV Transmission Group

BACKGROUND

The effect of sexually transmitted disease (STD) treatment on HIV transmission is a topic of considerable current interest and controversy.

GOAL

To assess the potential effect of STD treatment on HIV transmission in persons who are dually infected with STD and HIV.

STUDY DESIGN

Using data from eight STD clinic sites in the United States, the authors estimate the actual achievable reduction in HIV transmission by multiplying the prevented fraction associated with treatment of STDs (set at an average of 0.8) by the maximum potential reduction in HIV transmission achieved by treating STDs (using an average relative risk of 3.0 for increased HIV transmission in the presence of STDs). Subgroup analysis assessed infection proportions for genital ulcer disease, nonulcerative STDs, and any STD by sex, ethnicity, age, and sexual orientation.

RESULTS

The maximum achievable reduction in HIV transmission from dually infected persons to their partners is approximately 33%. The actual achievable reduction is approximately 27% (range, 10.0-38.1%) at the eight clinic sites. If each of the 4,516 dually infected persons in this cohort experienced a single sexual exposure with an uninfected person, 28 HIV infections would occur in the absence of STD treatment whereas 16 infections would occur with STD treatment.

CONCLUSIONS

The estimate of a 27% reduction provides a qualitative assessment of the potential impact of STD treatment on HIV transmission in the absence of any other behavioral intervention. Identification of dually infected persons in STD clinics is an important mechanism for targeting interventions to a social milieu with high risk for HIV infection and other STDs.

Chlamydia transmission: concurrency, reproduction number, and the epidemic trajectory

To identify factors that influence individual and group transmission of Chlamydia, the authors conducted community-wide contact tracing of chlamydia cases in Colorado Springs, Colorado, from mid-1996 to mid-1997. Case patients identified persons with whom they had had contact during the 6 months preceding diagnosis; contacts were actively sought and offered DNA amplification testing. Sexual contact networks were used to identify "source cases" and "spread cases," permitting estimation of the basic reproduction number (R0) for individuals and groups. Network and epidemiologic factors influencing R0 were assessed using univariate and multivariate procedures. Of 1,309 case patients, 1,131 (86%) were interviewed, and 2,409 contacts were identified. The 1,131 interviewed cases yielded 623.9 computed spread cases, for an overall R0 of 0.55. Few subgroups analyzed yielded a mean R0 exceeding unity-an observation in keeping with routine surveillance information which suggests that chlamydia incidence is declining in Colorado Springs. Concurrency, a network measure of simultaneous partnerships, was the most powerful predictor of transmission. Direct estimation of basic reproduction numbers for chlamydia using contact tracing techniques is feasible and can produce useful data with which to prioritize control efforts, evaluate interventions, and gauge the place of chlamydia on the epidemic continuum.

Invoking, monitoring, and relinquishing a public health power: the health hold order

OBJECTIVE

To describe a quarter-century's use of a public health power (Health Hold Orders) as an adjunct to noncoercive sexually transmitted disease (STD) control efforts in a middle-American city.

METHODS

Persons arrested for prostitution were involuntarily detained for up to 72 hours if they had not been tested for STD within 30 days of arrest. Such persons were mandatorily tested/treated for STD and voluntarily tested for HIV by health department providers in Colorado Springs from mid-1970 through 1994.

RESULTS

Prostitutes viewed temporary detention as inconvenient, but not as inappropriate. Over the 25-year interval, 4,965 examinations in prostitutes yielded 818 positive gonorrhea tests; the 1,564 tests performed under the health-hold order yielded 218 positive results. Positivity rates among prostitutes locally for reportable STD/HIV declined substantially during the period of observation, providing support for termination of the involuntary detention system.

CONCLUSIONS

The involuntary detention system contributed to observed communitywide declines in STD/HIV prevalence. Our experience demonstrates the importance of surveillance and empiric validation in public health practice.

Network structural dynamics and infectious disease propagation

We aimed to relate dynamic changes in risk-network (sex and/or injecting drug) structure to observe STD/HIV transmission. We analysed macro- and micro-structural elements in 2 heterosexual networks, augmented by ethnographic observations. In a Colorado cohort of injecting drug users (n = 595), measures of subgroup formation and of density of activity show decrease of network cohesion over time; only one HIV transmission was observed in 3 years. In a group of adolescent heterosexuals in Georgia (n = 99), the reverse process (increase in structural cohesion) was associated with efficient syphilis transmission: 10 cases were observed. Changes in personal risk behaviours over time were modest. STD/HIV transmission patterns were associated with intensification or diminution of network cohesion. Network and ethnographic data suggest that enhanced connectivity facilitates transmission while segmentation impedes it, suggesting opportunities for interventions. These data also emphasize the need to re-evaluate purely behavioural explanations of STD/HIV transmission.

Social network dynamics and HIV transmission

OBJECTIVE

To prospectively study changes in the social networks of persons at presumably high risk for HIV in a community with low prevalence and little endogenous transmission.

METHODS

From a cohort of 595 persons at high risk (prostitutes, injecting drug users, and sexual partners of these persons) and nearly 6000 identified contacts, we examined the social networks of a subset of 96 persons who were interviewed once per year for 3 years. We assessed their network configuration, network stability, and changes in risk configuration and risk behavior using epidemiologic and social network analysis, and visualization techniques.

RESULTS

Some significant decrease in personal risk-taking was documented during the course of the study, particularly with regard to needle-sharing. The size and number of connected components (groups that are completely connected) declined. Microstructures (small subgroups of persons that interact intensely) were either not present, or declined appreciably during the period of observation.

CONCLUSIONS

In this area of low prevalence, the lack of endogenous transmission of HIV may be related in part to the lack of a network structure that fosters active propagation, despite the continued presence of risky behaviors. Although the relative contribution of network structure and personal behavior cannot be ascertained from these data, the study suggests an important role for network configuration in the transmission dynamics of HIV.

Using social network and ethnographic tools to evaluate syphilis transmission

BACKGROUND AND OBJECTIVES

Partner notification has been the cornerstone for the prevention and control of syphilis in the United States. This technique may not make full use of contextual data that an ethnographic and social network approach can offer.

GOALS OF THE STUDY

The occasion of a syphilis outbreak among young people was used to investigate the applicability of a social network approach and to test the validity of several traditional approaches to syphilis epidemiology.

STUDY DESIGN

An outbreak of syphilis was investigated by interviewing both infected and noninfected people, by directing resources based on network association, by creating and evaluating network diagrams as an aid to the epidemiologic process, and by including ethnographic observations as part of outbreak management.

RESULTS

Diagrammatic display of network growth provided a useful alternative to the traditional epidemic curve. Case prevention was demonstrated by identifying uninfected people with multiple concurrent exposures. Concurrent, overlapping exposure in infected people rendered traditional "source" and "spread" criteria moot.

CONCLUSIONS

The current discussions of partner notification may be informed by recognizing that it is a subset of a broader and potentially more powerful approach. This approach calls some basic tenets of syphilis epidemiology into question.

Sociometric risk networks and risk for HIV infection

OBJECTIVES

This study examined whether networks of drug-injecting and sexual relationships among drug injectors are associated with individual human immunodeficiency virus (HIV) serostatus and with behavioral likelihood of future infection.

METHODS

A cross-sectional survey of 767 drug injectors in New York City was performed with chain-referral and linking procedures to measure large-scale (sociometric) risk networks. Graph-theoretic algebraic techniques were used to detect 92 connected components (drug injectors linked to each other directly or through others) and a 105-member 2-core within a large connected component of 230 members.

RESULTS

Drug injectors in the 2-core of the large component were more likely than others to be infected with HIV. Seronegative 2-core members engaged in a wide range of high-risk behaviors, including engaging in risk behaviors with infected drug injectors.

CONCLUSIONS

Sociometric risk networks seem to be pathways along which HIV travels in drug-injecting peer groups. The cores of large components can be centers of high-risk behaviors and can become pockets of HIV infection. Preventing HIV from reaching the cores of large components may be crucial in preventing widespread HIV epidemics.

Personal risk taking and the spread of disease: beyond core groups

Disease control efforts directed at human immunodeficiency virus are predicated on the need to reduce personal risk behaviors; that approach may not adequately reflect the complicated interplay between personal behaviors and the social setting in which they occur. Efforts to date, including the application of population ecology, the development of the core group hypothesis, and the use of compartment models to describe disease transmission, have aided in understanding the dynamics of transmission and have highlighted the relationship between personal risk taking and population risk. An area for further development is the application of the techniques of social network analysis to infectious disease spread. Initial work suggests that social structure may act as a barrier (or facilitator) in disease transmission and that the epidemiologic impact of a risky act varies with the social setting. The local context for risk behaviors has important implications for the dynamics of transmission.

Delivering breast and cervical cancer screening services to underserved women: Part II. Implications for policy

Many breast and cervical cancer screening (BCCS) programs for underserved women employ strategies to increase the use of preventive services. In Phase I of a two-phase study, strategies were identified and assessed. In Phase II, we further assess strategies previously identified and comment on policy implications. Site visits were conducted at BCCS programs that had used one successful strategy identified during Phase I, provided services to underserved women, and were located in different geographic regions. The federally funded National Breast and Cervical Cancer Early Detection Programs (NBCCEDP) were also considered for site visits. Interviews were completed and available data were reviewed. A descriptive and qualitative analysis was completed. Programs visited were found to be increasing the use of BCCS services for the defined target populations. Some programs focused on outreach and recruitment. Other programs focused on clinical preventive services with little emphasis on outreach and recruitment. Management information systems were used by most programs. We found that there continues to be a large number of women not receiving BCCS services. Some programs have had to limit outreach and recruitment because the clinical preventive services offered are at capacity. Programs need to have a balanced approach to providing services from the outset. Existing programs may need to establish partnerships to provide comprehensive BCCS services to underserved women. Because the unmet need (women who have not received BCCS services) exceeds available clinical preventive services, it is a challenge to know how to best use available resources.

Delivering breast and cervical cancer screening services to underserved women: Part I. Literature review and telephone survey

Although breast and cervical cancer screening procedures have been shown to reduce morbidity and mortality, many women are not using these services. These women are likely to be older, of ethnic or racial minority, of low socioeconomic status, less educated, underinsured, or living in rural locations. Many breast and cervical cancer screening programs employ strategies to increase use. In order to identify and assess those strategies, we reviewed the literature and completed a telephone survey, altogether assessing 61 programs. Our study identified several strategies and found that not all strategies work for all women. Management systems directed to both patients and providers consistently are effective for most underserved women. Community-based outreach and integration of preventive services at the primary health care (PHC) site are effective strategies for both African American and Hispanic women. Use of mass media has been successful when targeted toward Hispanic women, but not when targeted toward African American women. Mobile units and integration of preventive services at PHC sites are effective strategies for elderly women. In many cases, a combination of strategies may be effective. Programs should employ strategies best suited to their target populations.

Syphilis control. The historic context and epidemiologic basis for interrupting sexual transmission of Treponema pallidum

Syphilis control has been the prototypic sexually transmitted disease (STD) public health program of the 20th century. However, the disease remains nearly as much an epidemiologic enigma as it did in the early 1900s. This article examines the historic and epidemiologic bases for syphilis control, using unpublished data to supplement a recent model of STD transmission. The authors recommend building on such traditional individually oriented strategies as case finding, partner notification, and presumptive treatment as a basis for future community-oriented, population-based strategies including (but not limited to) selective mass treatment in high-prevalence populations. Using epidemiologic information to target population-level interventions will be the paradigm for syphilis control in the 20th century.

The natural history of exercise: a 10-yr follow-up of a cohort of runners

To determine the pattern of exercise and associated adverse events, including injuries, a mail survey was conducted on participants in the 1980 Peachtree Road Race Study for whom we were able to obtain current addresses. A total of 535 persons responded, 326 men and 209 women, representing 72% of those with identifiable addresses. Although only 56% of respondents reported that they were still running in 1990, 81% reported that they were still exercising regularly. The cumulative probability for continuing to run was 0.71 for men and 0.56 for women. Injury (31%) was the chief reason given by men for stopping permanently. For women, 28% stopped because they chose another form of exercise. Fifty-three percent of respondents had at least one injury during the 10-yr interval. The probability of experiencing an injury was associated with higher weekly mileage. The knee was the most frequently injured site. Thirty-nine percent of women and 35% of men reported being verbally assaulted. Approximately 10% of persons were hit by thrown objects or bitten by a dog. In a group of recreational runners, almost half had stopped running 10 yr later, but over 80% were still physically active. Many runners sustain injuries or suffer hazards related to their exercise.

Public health surveillance for chronic conditions: a scientific basis for decisions

In this paper we investigate the important contribution of multiple public health surveillance systems to policy in chronic disease control and prevention. We show that, typically, surveillance for chronic diseases relies on multiple data sources, often created for another purpose. We also define the concept of burden for chronic conditions based on data from multiple sources. An example from a state illustrates a model for combining data for use in policy development. These applications illustrate the central role of statistical methods in ensuring the appropriate use of data from multiple surveillance systems.

Confidence intervals, hypothesis tests, and sample sizes for the prevented fraction in cross-sectional studies

The prevented fraction (PF) is the proportion of disease occurrence in a population averted due to a protective risk factor or public health intervention. The PF is not equivalent to the population attributable risk (AR). The AR is appropriate for epidemiologic studies of disease etiology, and for estimating the potential impact of modifying risk factor prevalence. The PF more directly measures the impact of public health interventions, however, and thus is an important evaluation tool. We derived the variance of the estimated PF by using maximum likelihood theory for cross-sectional studies. We used simulations to compare the performance of confidence intervals based on various transformations of the estimated PF. The logit transformation was the best choice when PF > or = 0.3, whereas the untransformed estimate was best when PF < 0.3. We present formulae for hypothesis testing and sample size calculations, discuss the issues of interaction and confounding and give two estimators adjusted for confounding.

Mapping a social network of heterosexuals at high risk for HIV infection

OBJECTIVE

To determine how heterosexuals at risk for HIV infection interconnect in social networks and how such relationships affect HIV transmission.

DESIGN

Cross-sectional study with face-to-face interviews to ascertain sociosexual connections; serologic testing.

PARTICIPANTS

Prostitute women (n = 133), their paying (n = 129) and non-paying (n = 47) male partners; injecting drug users (n = 200) and their sex partners (n = 41). Participants were recruited in sexually transmitted disease and methadone clinics, an HIV-testing site, and through street outreach in Colorado Springs, Colorado, USA.

MAIN OUTCOME MEASURES

Reported behaviors, risk perceptions, sociosexual linkages, and HIV prevalence.

RESULTS

Respondents were well informed, but reported engaging in high-risk behaviors frequently. Nevertheless, over 70% of respondents perceived themselves to be at low risk for HIV infection. The 595 respondents identified a social network of 5162 people to which they belonged. Network analytic methods indicated 147 separate connected components of this network; eight of the 19 HIV-positive individuals in the network were located in smaller components remote from the largest connected component.

CONCLUSION

The isolated position of HIV-positive individuals may serve as a barrier to HIV transmission and may account for the lack of diffusion of HIV in heterosexual populations in this region. Network analysis appears useful for understanding the dynamics of disease transmission and warrants further development as a tool for intervention and control.

Competing mortality and progress against cancer

The decline in ischemic heart disease (IHD) mortality in the United States in recent years is thought to have contributed to increases in cancer mortality. To estimate the interrelation between these competing causes of death between 1970 and 1988, I constructed a hypothetical population schedule by assuming that age-specific IHD mortality risks had not declined. The difference between the actual population and the hypothetical population represents persons who did not die from IHD and were thus available to die from cancer. Using observed age-specific cancer risks over the entire interval, 153,207 of the 7,649,058 cancer deaths (2.0%) in persons age 20-85 years occurred in IHD survivors; in 1988, 24,053 of the 482,490 cancer deaths (5.0%) occurred in IHD survivors. Among 55 to 85 year olds in 1988, IHD survivors accounted for 5.5% of the cancer deaths. Alternative assumptions about the susceptibility of IHD survivors to cancer have little impact on the contribution of IHD survivors to cancer deaths. Results from a separate analysis demonstrated that the proportional contribution of true cancer risk to the increase in cancer cases tripled in the interval 1970-1988 compared with the interval 1930-1970. These observations indicate that the contribution of the IHD mortality decline to the increase in cancer mortality has been small and does not account for the increasing age-specific risks for cancer among older persons.

Confounding in community interventions

Evaluations of community intervention projects usually use the community as the unit of analysis and employ a possibly costly cohort approach, which may not achieve sufficient power. The case-control approach, applied extensively in cancer control, may be an inexpensive alternative. Moreover, the case-control approach permits evaluation of potential confounding by factors possibly associated with the intervention. A simulation based on case-control methodology for calculating an intervention program effect (prevented fraction) demonstrates that the threat of confounding, at levels of prevented fractions in the range of 10% to 30%, is modest. Community-based interventions with lower prevented fractions are more susceptible, and their results should be considered with caution. Such boundaries may be of use to investigators in planning and evaluating community-based efforts to reduce risk and control disease.

AIDS in Colorado Springs: is there an epidemic?

OBJECTIVE

To analyze trends and patterns of HIV infection in a medium-sized community in the United States.

METHODS

Surveillance for AIDS and HIV infection was conducted by private physicians, military and public clinics, and blood and plasma donation centers. HIV-positive individuals were contacted and asked to refer their sex and injection partners for HIV-antibody testing. Prostitutes, injecting drug users and their sex partners were studied. Selected physicians were surveyed to assess under-reporting.

RESULTS

The 740 HIV-infected adults (67 with documented seroconversion) included 506 with no evidence of AIDS, 58 living with AIDS, and 176 who had died. Of the 126 patients cared for by local physicians, 107 (85%) had been reported. No major changes in behavioral risk factors or increases in the number of HIV-infected individuals occurred between 1986 (128) and 1992 (95).

CONCLUSIONS

Characteristics of individuals at risk and incidence of HIV infection have remained stable from 1981 to 1992. Analysis of data from the comprehensive surveillance and control program established in Colorado Springs in response to the AIDS epidemic suggests that, unlike the nation's epicenters, HIV incidence in this location is neither widespread nor rapidly increasing. The age distribution of reported cases is slowly increasing, and the ratio of newly reported cases to deaths is declining, implying stable or decreasing incidence; deaths may soon exceed new cases. Using data routinely available to public health officials, we conclude that the epidemiologic picture of AIDS--like the clinical one--must be heterogeneous, and that rational planning for the impact of AIDS should be based on the collection and analysis of local data.

Excess deaths from nine chronic diseases in the United States, 1986

To assess excess mortality from chronic disease in the United States, state age-adjusted combined mortality rates for nine chronic diseases in 1986 were compared with three "minimum" rates--two calculated from rates actually achieved in states and a third estimated as the mortality remaining after elimination of one risk factor for each disease. Hawaii had the lowest mortality rate of combined diseases (305/100,000); state excesses ranged from 0% to 37%. The sum of lowest disease-specific rates in any state was 284 per 100,000, indicating excesses of between 7% and 41%. A minimum mortality rate of 224 per 100,000 was estimated to result from elimination of one risk factor for each of the nine diseases, indicating state excesses from 26% to 54%, or 524,000 US deaths. Reduction of US mortality from the nine diseases to the risk factor--eliminated rate is estimated to be associated with an increased life expectancy at birth of 4 years.

Application of the Cox model as a predictor of relative risk of coronary heart disease in the Albany Study

Patients in long term studies of coronary heart disease may have different levels of risk during the course of study. Smoking habits, blood pressure, and obesity may change drastically during this period. The multiple logistic model, the most commonly used model for the analysis of coronary heart disease studies, does not consider survival time in assessment of the dependent covariates and does not account for the censoring which usually occurs in such studies. We propose a Cox model with time-dependent covariates to model the risk of coronary heart disease in the Albany study. The Cox model we fitted evaluates the patients' risk on the basis of the data at the last visit. With this methodology, we can evaluate whether it is advantageous for individuals to modify their risk of disease by their effecting changes in their covariates, that is to stop smoking, lose weight, change diet and so on. The important covariates that explain the risk of coronary heart disease were the same in our model as in the models used in the earlier reports. The estimated relative risks were slightly higher in most cases and lend more support to the need to encourage patients to achieve a better covariate state.

Ischemic heart disease and hypertension: effect of disease coding on epidemiologic assessment

During the changeover from the eighth to the ninth revision of the "International Classification of Diseases: Manual of the International Statistical Classification of Diseases, Injuries, and Causes of Death" (ICD), there were several major alterations of coding for the rubrics Ischemic Heart Disease (IHD) and Hypertension (HBP). As expected, these changes caused major discontinuities for IHD and HPB. These discontinuities were not, however, uniform over sex-race groups. When examined by component ICD codes, the discontinuities were found to vary in both magnitude and direction among the groups. In addition to discontinuity, there was a change in the rate of decline for IHD and HBP after the changeover. This rate of decline varied as well by sex-race group. In general, the decline among blacks was slower than among whites. Earlier studies that assessed IHD mortality have used different groupings of ICD codes to obviate the discontinuity, and researchers have observed a similar differential decline. These results should be viewed with caution because of the potential impact of differential coding on sex-race groups. As preparations are made for ICD-10, special attention should be given to the preservation of epidemiologic continuity to provide better assessment of trends in population subgroups.

Epilogue. Establishment of the Center for Chronic Disease Prevention and Health Promotion

The establishment of the Center for Chronic Disease Prevention and Health Promotion (CCDPHP) at the Centers for Disease Control (CDC) following the Conferences on the State of the Art in Quality Control Measures for Diagnostic Cytology Laboratories is briefly discussed. The CCDPHP is expected to play a major role in the CDC's cancer control program, including participation in establishing effective screening programs and assuring the quality of such methodologies as the Papanicolaou test and mammography.

Identifying important results from multiple statistical tests

When many statistical tests are performed simultaneously, the overall chance of a type I error (incorrect rejection of a true null hypothesis) can substantially exceed the nominal error rate used in each individual test. Numerous techniques exist to adjust results of individual tests to control this problem. In general, these techniques apply a more stringent criterion of statistical significance (a smaller P-value) to each individual test than normally needed to maintain the experimentwise type I error. With an analysis that seeks to identify results for further research, however, such a conservative technique may not be appropriate. We present a new approach that uses a mixture of several distributions to model the set of P-values or of test statistics. One component models the results consistent with a failure to reject the null hypothesis, while the other distribution(s) in the mixture represent results inconsistent with the null hypothesis. These latter results may not achieve statistical significance based on a conventional P-value. We illustrate the use of the method on national mortality data and on several data sets analysed previously.

Temporal and social aspects of gonorrhea transmission: the force of infectivity

Theoretical consideration of the epidemiology of sexually transmitted diseases (STD) suggests that the key determinants of transmission dynamics are the duration of infectiousness and the extent to which subgroups in the population interact sexually. We used two empirical correlates to represent these concepts: (1) the force of infectivity, calculated by summing all the days of potential infectivity (the time between last sexual exposure and treatment of the contact) generated by a given case, and then summing the days for all cases within a given subgroup; (2) self-selection, representing the observed probability that members of a given subgroup select sexual partners from within their own group. Using data gathered in Colorado Springs, Colorado, we estimated that a single group i.e., black, male, heterosexual, military personnel residing in the core areas, generated 27% of the force of infectivity. Subgroups that select greater than 50% of their sexual partners from outside their sociodemographic boundaries generated a rate for the force of infectivity that was 4.5 times higher than the rate for self-selectors. These findings confirm the core group theory and suggest strategies for control of STD.

Disease transmission by heterosexual men with gonorrhea: an empiric estimate

In Colorado Springs, Colorado, between January and June 1981, approximately 400 cases of gonorrhea in heterosexual men generated approximately 200 "spread" cases in women. The ratio of spread cases to index cases (0.5) is an empiric measure of the reproductive rate for the disease, i.e., the rate at which an infected person replaces himself. Although a reproductive rate of less than one theoretically should produce extinction of the disease, the rate derived from these data is shown to be a weighted average of the rates in symptomatic men (0.31), subsymptomatic men (0.79), and asymptomatic men (1.30). Approximately 35% of gonorrhea transmitted by men may be attributed to the asymptomatic group, which constitutes 10-15% of the infected population. If the observations in this community are applied to the nation as a whole, we estimate that approximately 1.7 million cases of gonorrhea occurred in 1981, as compared with the 990,864 cases actually occurred in 1981, as compared with the 990,864 cases actually reported. This estimate is close to those usually quoted but differs in gender composition. Whereas the observed male-to-female ratio is 1.48:1, the predicted ratio, based on estimates of reproductive rates in subgroups, is 0.8. This analysis supports the theoretical notion that the endemicity of gonorrhea is maintained by heightened transmission within small groups, and it suggests that underdetection is more important than underreporting in the assessment of the "true" incidence of this infection.

A controlled trial of disease surveillance strategies

Active surveillance techniques using routine telephone contacts with providers improved the reporting of measles, rubella, salmonellosis, and hepatitis by a factor of 4.6 among private physicians in Monroe County, New York, and increased reporting for these target diseases from all sources by 51 percent. The timeliness of reporting was not improved by active surveillance. Reporting patterns varied by disease and source of report, suggesting the desirability of various approaches to surveillance based on local resources and priorities. Although reporting rates were higher for diseases among persons from census tracts of low socioeconomic status, physicians providing care to persons living in low-income areas responded no differently to active reporting than did those providing care to patients from middle- and high-income areas.

The community ecology of diabetes patient classification and practice characteristics

We examined the clinical and epidemiologic characteristics of diabetic subjects ascertained from five different sources (hospital, private practice, specialist practice, nursing home, and prevalence survey) in a medium-sized upstate New York community. Patients were categorized using the National Diabetes Data Group classification (based on weight and insulin requirement) and using an alternative clinical classification scheme based on age at onset and insulin use. Both classifications demonstrate marked differences in the distribution of diabetic subjects. Those seen in general medical practice reflect the prevalent population and tend to have fewer complications and less insulin requirement. Those seen in specialist practice or in hospital have higher complication frequencies. Thus, the source of patients may determine the epidemiologic picture of the disease, as well as the practitioners' perceptions. Use of the clinical classification scheme helps to identify a subgroup of diabetic subjects--those with onset past the age of 25 yr who use insulin--at greatest risk of complications. Further investigation of high-risk groups is warranted, particularly for the implementation and evaluation of diabetes control programs.

Observations on the application of EPI cluster survey methods for estimating disease incidence

The present study attempted to assess the incidence of target diseases of the Expanded Programme on Immunization (poliomyelitis, tetanus, measles, pertussis, neonatal tetanus, diphtheria), using cluster samples and a household interview form. The results suggest that this method can indeed serve to estimate the incidence of these diseases with reasonable precision and may also be used to demonstrate reduction in incidence for the more common diseases. Analysis of 37 surveys for poliomyelitis and neonatal tetanus in India revealed a relative uniformity in the design effect (i.e., the ratio of the variance for the cluster estimate to the variance for the binomial estimate) for diseases with low incidence and prevalence. Diseases with higher prevalence tend to have a larger design effect, which may be indicative of the epidemic and "clustered" nature of the disease. A large design effect, therefore, does not necessarily indicate a need for a larger sample size, particularly if precision is acceptable. There is no one single design that is ideal for all surveys of disease incidence and decisions must be made in the light of local conditions and available resources.

Gonorrhea as a social disease

Gonococcal infection in Colorado Springs, Colorado, is concentrated in about 1% of the population. The social groups at risk are characterized as young, nonwhite, heterosexual, and connected to the military. They exhibit residential proximity by clustering in "core" census tracts; 51% of cases were in four tracts. They demonstrate residential stability and close social association at preferred sites for nighttime leisure activity (six major sites out of 300 available). Social aggregation is further demonstrated by the length of social contact prior to sexual contact (45% had known each other for over two months), the neighborhood nature of sexual choices, and the grouping of sexually connected individuals in lots (six lots contained 20% of cases). The force of infectivity, measured in person-days of potential spread of gonorrhea by infected contacts, provides a quantitative assessment of the important of identifiable social groups in the transmission of gonorrhea.

A civilian-military partnership to reduce the incidence of gonorrhea

To reduce the incidence of gonorrhea in the Colorado Springs, Colo., area, casefinding measures (interviewing of patients and tracing of contacts) were conscientiously applied by the local health department, in cooperation with the U.S. Army, to more than 90 percent of reported cases during a 3-year period. Nearly 4,000 cases of gonorrhea--2,127 civilian and 1,811 military--were interviewed; they named 7,399 contacts. A total of 1,141 cases of gonorrhea were newly identified and patients brought to treatment in Colorado Springs as a result. Implementation of these measures was associated with a 12.9 percent overall decline in gonorrhea incidence. This decline was most pronounced in the civilian population (20 percent), while little change in incidence occurred in the military population. The data presented suggest that the orderly application of casefinding epidemiology, allied with other control program initiatives, can interrupt transmission of, and prevent, disease.

The geography of gonorrhea. Empirical demonstration of core group transmission

The pattern of reported gonorrhea in Upstate New York (exclusive of New York City) in the years 1975-1980 is one of intense central urban concentration, with concentric circles of diminishing incidence. The relative risk for gonorrhea in these central core areas, compared to background state rates, is 19.8 for men and 15.9 for women, but as high as 40 in selected census tracts. Prevalence appears to approach 20% in some areas, the level postulated by current epidemiologic models for continuing endemic transmission. These core areas are characterized by high population density, low socioeconomic status and a male to female case ratio of one or lower. Contact investigation data suggest that sexual contact tends to exhibit geographic clustering as well. These observations provide support for narrow focusing of epidemiologic resources as a major disease control strategy.

Gonococcal pelvic inflammatory disease: case-finding observations

During a 20 month period, 110 women with gonococcal pelvic inflammatory disease an 165 women with uncomplicated gonorrhea were provided intensive case-finding services (interviewing of patients and tracing of contacts). Approximately three contacts per case were investigated, and 24.3% of the 859 male contacts were infected. Nearly two thirds (64.6%) of the infected contacts were asymptomatic. Active public health intervention was frequently necessary to persuade asymptomatic men to seek medical attention; removal of these men from the disease pool may serve to prevent reinfection of treated women and to diminish the transmission of gonorrhea.

Focused interviewing in gonorrhea control

To develop an operational approach to the identification of high risk gonorrhea transmitters, three groups of women infected with Neisseria gonorrhoeae (recent repeaters, routine discoveries, and women with pelvic inflammatory disease) were offered intensive casefinding services during an 18-month period. Approximately three contacts per case were investigated, and 27.4% of the contacts were infected. Of infected contracts, 61% were asymptomatic. Asymptomatic, remote contacts to these women appear to be important in the continuing transmission of gonorrhea. The interviewing approach used reflected that employed in syphilis (thorough, detailed, and long) rather than the more casual interviews usually employed for gonorrhea patients. During this period, gonorrhea morbidity declined 22%. Further exploration of a targeted approach to gonorrhea epidemiology is indicated.

Analysis of routine data describing morbidity from gonorrhea

A system was developed for the collection, coding, and processing of data from routine reports describing morbidity caused by gonorrhea. These data are coded centrally, and tables that provide information about the age, sex, race, source of care, and geographic and temporal distributions of the population affected by gonorrhea are generated. Applications of the system are described; these include study of distribution of the disease, trends in utilization of facilities, seasonal trends in incidence of the disease, and several features of recidivism in the Denver (Colorado) metropolitan area. This system has potential as a tool for evaluation of disease control efforts in any standard metropolitan statistical area.

Strategic planning system for control of venereal disease: record keeping in a clinic for treatment of sexually transmitted diseases

A manual data processing system and a computer-based system that was integrated into the operations of a clinic for treatment of sexually transmitted diseases were developed. The automated system can generate routine management reports and cross-tabulate data for special studies. It is flexible and adaptable to other localities but is not a panacea for clinical problems.

Efficacy of selected diagnostic tests for sexually transmitted diseases

Direct urethral and cervical smears for gonorrhea in women are useful because of their high specificity. Male rectal smears are less efficacious. Direct tests for candidiasis and trichomoniasis are also specific, but less sensitive than generally thought. Most tests exhibit only moderate month-to-month variability. No association between infection with trichomonads and gonococci was demonstrated. The gonococcal tests all perform best under conditions of high prevalence, such as those that obtain in venereal disease clinics in the United States.