Sexually Transmitted Diseases  >  Program Guidelines  >  Surveillance and Data Management

Program Operations Guidelines for STD Prevention
Surveillance and Data Management

Previous Page

Table of Contents

Next Page

Sections on this page:

• DATA ANALYSIS, INTERPRETATION, AND DISSEMINATION
• Planning an Analysis
• Performing the Analysis
• Interpretation of data
• Development and evaluation of screening criteria
• Dissemination and Communication

DATA ANALYSIS, INTERPRETATION, AND DISSEMINATION

Data analysis and interpretation are necessary to accomplish the purposes of case-reporting and prevalence monitoring. This section focuses on those types of analyses that should routinely be performed by STD prevention programs and by the Regional Infertility Prevention Projects. The analyses recommended here are intended to assist program efforts to properly address STD surveillance activities. Data analysis required for evaluating and assuring data quality is discussed later in this chapter.

Planning an Analysis

The collection, analysis, and dissemination of local STD data are acknowledged as important components of the efforts to prevent sexually transmitted disease. STD project areas continue to collect large amounts of data. Unfortunately, all too often, the only apparent reason for collecting these data is "to satisfy CDC reporting requirements" that enable the assessment of national STD morbidity levels and characterization of persons infected with STDs. Although important, this reason is usually not relevant when considering state and local-area informational needs.

The reality is that the sheer number and variety of demands placed on program managers very often leaves little time and energy for analysis. This can result in a "catch 22" for managers—little time for the analysis that can help to answer the questions that are vitally important to the effective operation of STD prevention programs at the state and local levels. Who are the infected persons? Who is the at-risk population? Where is transmission occurring? Which interventions work best with different groups of infected persons and when and how should they be implemented? How effective was the targeted intervention in reducing risk or infections? Are we doing a good job of controlling STD in the community, and how do we know it? What are the data needs of the community to assist in planning and policy efforts? These questions should be driving data collection and analysis efforts. Issues to consider in planning an analysis include:

• Proceed from the simple to the complex

Do not begin your analysis by trying to examine trends over time for four different STDs by various characteristics. Start by asking a question or stating a hypothesis that you want to answer. How many cases of a given STD were reported each year? Then look at characteristics of the case. Again, start with the simple ideas. How many cases of a particular STD were reported in each age group each year? What were the sex- specific rates? Only after looking separately at each characteristic should you begin looking at relationships between these characteristics. For example, how many cases in teenage males were reported? What was the rate for teen-age males? Proceed from the simple to the complex. For example, using EpiInfo or STD*MIS, run frequencies by data element in a given patient population.

• Know your data

Some of us may want to know immediately about the trends over time for several different STDs by various characteristics of the cases. However, what happens if our rates for the STDs by the characteristics of the cases are not consistent with our day-to-day observations about the STD problem in our community and in other communities similar to ours? Is it because our day-to-day observations are incorrect, or is it because there are problems with one or two cases aged 75 years? How did these two cases affect the computation of the average age of our cases? Only by knowing your data can you understand and appropriately interpret the results of more complicated analyses. There is no substitute for knowing your data. Individuals from the community, medical providers, outreach workers, etc., should be included in preparing to conduct analysis of the data. These individuals may ask questions of the data not typically asked in a STD prevention program. Asking the proper questions of the data can often drive the type and direction of the analysis needed.

• Ensure valid data

The final guiding principle of data analysis is a familiar one: "garbage in; garbage out." How much care is given to collecting and managing the data in standard ways so as to minimize error and maximize validity? Just because data are stored in a computer and because computer programs can generate official-looking reports from those data does not necessarily mean the data are valid. STD prevention programs must expend the additional time and effort necessary to ensure that STD data on which analysis can be performed are valid. This can be accomplished by reviewing source documents such as patient charts, laboratory reports, and case reports and comparing them to the final outcome data.

Performing the Analysis

Statisticians and epidemiologists receive training in a variety of analytical methods that require knowledge of the underlying statistical and mathematical foundations used to develop those techniques, and their proper application. Providing an exhaustive and detailed description of the different methods available to explore, summarize, analyze, or display surveillance data goes well beyond the scope of guidance that can be provided within this document. However, one reference that STD prevention program areas will likely find helpful in preparing for the analysis and interpretation of surveillance data is Principles and Practice of Public Health Surveillance (Teutsch, 1994). It presents an excellent overview of the types of analytical presentations and methods that are commonly employed, and it discusses many of the associated data issues that need to be considered in analyzing and interpreting data. In addition, examples of standard tabular and graphical data displays can also be found in any of the annual STD Surveillance Reports produced by CDC's Division of STD Prevention.

One very important general recommendation for programs is to have access to a statistician or epidemiologist. The STD prevention program staff and management should work with the epidemiologist to clarify what data need to be collected to answer the question. Without the availability of staff adequately trained and skilled in collecting valid data to answer proposed questions and in analyzing and interpreting analytical methods, the likelihood of misinterpretation and under-utilization of data increases, and therefore programs have a weakened ability to quantitatively monitor and describe the effects of STDs in their area.

Ideally, any analysis, summarization, graphical display, or interpretation of data should be based on data that are deemed reliable (i.e., reproducible), valid (i.e., accurate), complete, and timely. Statistical summaries based on data that do not satisfy these characteristics require a discussion of these shortcomings or limitations and the possible effect on the analysis and interpretation of the data being analyzed. Case counts and crude rates (i.e., unadjusted rates) are common measures of disease burden that often are presented with respect to time, by geographic area (e.g., county, zip code, or census tract) , personal characteristics (i.e., sex, age, race or ethnicity), and various combinations (e.g., age-by-race-by-sex). Crude rates are a measure of the actual public health burden from STDs in a community, and are appropriate to use for public health planning, policy making, and resource allocation.

Interpretation of Data

State and local STD prevention programs should perform the following analyses of STD surveillance data to facilitate monitoring of disease burden and trends.

Quarterly analysis

• Compare number of case reports, rates, and prevalence (if available) with same quarter during the previous year.
• Examine trend in the number of reported cases, rates, and prevalence for the past 1-5 years, overall, and by specific variables. For example, 1) geographic area, 2) sex, 3) age-grouping, 4) race or ethnicity, 5) provider-type, 6) provider-site, 7) laboratory, 8) test-type information, and 9) anatomic site might provide helpful insights. Those programs providing behavioral interventions, community health education, provider training, and partner services around selected STDs may consider additional categories such as reason for examination and individual risk factors.

Annual analysis

• Stratify case reports annually, by the nine variables listed above.
• Identify annual trend in overall population-based rates of reported cases, using the most recent census data, or the most recent intercensal estimates, and stratify by the basic demographic categories (geographic area, sex, age group, and race or ethnicity).
• Compare annual or point prevalence with prevalence for several previous years, overall, and stratified by the nine variables listed above.
• Examine rates of gonococcal or chlamydial coinfection.
• For clinic populations tested for HIV and other STDs, including chlamydia and gonorrhea, examine rates of HIV/STD co-infection. The following points should be considered when analyzing and interpreting STD data.
• Acute changes in reported morbidity or prevalence may be real or may be the result of changes in screening or reporting practices or test type.
• Case-reporting data can most reliably be used to monitor trends in disease burden when the amount of screening is stable, when no changes are occurring in access to and use of clinical services, and when diagnostic and reporting practices are consistent over time. A decline in the number of reported cases can result from a reduction in screening or from a decline in reporting by providers or laboratories.
• For comparing the disease burden of different groups, it is necessary to identify and use appropriate denominators. Census data stratified by geographic location, sex, age group, and race or ethnicity must be used for calculation of population-based rates of case reports for each of these variables.
• In all large jurisdictions, and especially in areas of high or increasing rates of disease, it is especially important to stratify simultaneously by sex, age, and race or ethnicity within each geographic area to determine which subgroups have the highest reported rates of disease.
• High prevalence of disease in an area where case reports are few suggests that screening coverage is inadequate or that cases are not being reported.
• Examination of number of case reports can be used to identify new sites that have recently begun more testing and reporting sites screening higher risk clients or those sites where there has been an acute decline in reported cases.
• Analysis between STD data sets and other data sets such as HIV, TB, Drug and Alcohol, etc., need to be conducted to determine co-infection rates, commonality of risk factors, and opportunities for collaboration.
• Behavioral and social surveillance should be a component of a comprehensive STD surveillance system.
• When stratifying data, confidentiality of individuals must be protected.

Development and Evaluation of Screening Criteria

Periodically, state and local STD prevention programs should work with participating clinics to evaluate their chlamydia and gonorrhea screening criteria, as these have implications for both case-reporting and prevalence-monitoring activities. To evaluate the sensitivity and positive predictive value of the provider's selective screening criteria, the STD prevention program can compare universal screening data with selective screening data initially, and periodically thereafter. Screening coverage according to screening criteria used should be evaluated periodically and at a minimum annually.

Dissemination and Communication

Feedback on data collection and analysis can take place at multiple levels. Careful monitoring of data for completeness and validity must be a regular part of data collection and interpretation. Inconsistencies in data collection, missing data, and other issues require immediate attention to ensure that reports provide information that accurately reflects program efforts.

Each state, local, and regional program should develop a plan to effectively communicate the analysis and interpretation of STD case-reporting and prevalence monitoring data to the general public, priority health care providers (especially those providing data), laboratories, community clinicians, support agencies, community-based organizations, HIV program directors, HIV community planning groups, corrections facilities, drug treatment centers, policy makers, public relations office, and other local, state, and national health care and public health agencies and partners. With the assistance of communications specialists and input from partners receiving the reports, state and local STD prevention programs should carefully tailor the communication of STD surveillance data to the specific needs of the target audience.

• Reports presenting case-reporting or prevalence monitoring data should include a concise interpretation of the data presented.
• When communicating the findings from their case reporting or prevalence monitoring activity, STD prevention programs should consider preparing the following types of reports:
  • annual report of case rates and prevalence by demographic variables and by trends;
  • reports to providers of data that identify their specific contributions to STD prevention efforts, prevalence data to guide their screening practices, and feedback on screening coverage;
  • newsletters or bulletins that provide clear, concise data interpretation and advice to clinicians, and laboratory directors, and community groups;
  • letters or limited reports to targeted providers that present information that would serve to stimulate additional reporting, e.g., reports of outbreak investigations or unexpected changes in epidemiological patterns;
  • limited reports (fact sheets for ready distribution for ad hoc requests); and
  • web page and two-way feedback to providers and laboratories.

When analyzed and organized for dissemination, selected STD case-reporting and prevalence monitoring data have been successfully used by many organizations to demonstrate the need for STD prevention services and to direct resources to specific populations. Studies have shown that contact between providers and health departments increased reporting; therefore, dissemination of useful STD surveillance to providers likely serves to increase or maintain reporting. State and local STD prevention programs garner good will from students, faculty, and other organizations that need data for grant and report writing by providing timely STD surveillance data. In some program areas, acknowledging the reporters of STD surveillance data in public reports has gained their good will and stimulated others to report.

Previous Page

Table of Contents

Next Page