How To Write A Quality Abstract For Submission To Unite For Sight's Global Health & Innovation Conference

Abstracts must demonstrate results-oriented research, and data is therefore crucial. Research is a vehicle for creating knowledge. Drawing incorrect conclusions from research, particularly in health-related studies, can have an impact on findings themselves, but more importantly on the implications and recommendations for improving patient care or community programs.

What is an example of a low-quality abstract that is without data?

Example of a low-quality abstract: HIV prevalence is high among women in Country A, but there is little research about HIV/AIDS perceptions in the region, and there are few educational programs about HIV. This study involved in-depth interviews with 150 women in each of three villages in Country A, followed by an educational intervention and post-intervention interviews. To be eligible, the women had to be married and at least eighteen years old. Prior to the educational intervention, the women reported awareness about HIV, and most correctly identified that HIV is sexually transmitted. The analysis demonstrated statistically significant improvements in knowledge about preventing HIV after the educational intervention. The post-intervention interviews revealed that 82% of participants felt that the intervention was beneficial. These results suggest that educating women is an effective way of increasing awareness about and reducing the prevalence of HIV.

Why is this abstract considered low quality?

• Sentence 1 (HIV prevalence is high among women in Country A, but there is little research about HIV/AIDS perceptions in the region, and there are few educational programs about HIV.): This sentence is acceptable, though broad statements are included without supporting facts.

• Sentences 2 and 3 (This study involved in-depth interviews with 150 women in each of three villages in Country A, followed by an educational intervention. To be eligible, the women had to be married and at least eighteen years old.): These sentences are intended to describe the research methodology. However, how were the women recruited for the study? Was an announcement made in the villages to invite women to attend an educational program and to participate in a research study? If so, there may be a sampling bias. The women self-selected themselves and chose to participate in the educational program. They may have a higher educational level than others in the village. Or, they may be more familiar with HIV and therefore had chosen to participate in the program. It is important to know how the women were selected to participate since this has an impact on the results.

• Sentence 4 (Prior to the educational intervention, the women reported awareness about HIV, and most correctly identified that HIV is sexually transmitted.): This sentence is unacceptable. How many of the 150 women reported awareness about HIV? How are the researchers defining and assessing "awareness"? How many of the 150 women correctly identified that HIV is sexually transmitted? The researcher states that "most" of the women correctly identified that HIV is sexually transmitted, but this is not defined by any numerical values.

• Sentence 5 (The analysis demonstrated statistically significant improvements in knowledge about preventing HIV after the educational intervention.): This sentence is unacceptable for the abstract. First, what type of educational intervention was implemented? The researcher also states that the analysis demonstrated "statistically significant improvements" in knowledge after the educational intervention. However, what data has been utilized to draw this conclusion? What survey questions were asked of the participants prior to and after the educational intervention? What were the common responses prior to and after the educational intervention, and what is the associated data?

• Sentence 6 (The post-intervention interviews revealed that 82% of participants felt that the intervention was beneficial.): This sentence is unacceptable for a quality abstract. What does “beneficial” actually entail? How was it measured, and are we sure that all women evaluated benefit in the same manner? Surveys evaluating the outcomes of an intervention must use language that is specific and pertains to a particular variable; subjective questions that ask participants to rate an intervention’s overall usefulness will not yield quality information about the actual impact and outcomes of a program.

• Sentence 7 (These results suggest that educating women is an effective way of increasing awareness about and reducing the prevalence of HIV.): This sentence is unacceptable. A conclusion is provided, though the abstract does not provide data to support this conclusion.

How should the abstract be revised?

According to the World Health Organization, HIV prevalence is ___% among women in Country A. This study involved in-depth interviews with 150 women in each of three villages in Country A who were married, at least eighteen years old, and enrolled in self-help groups. Each woman participated in an in-depth interview to assess understandings and perceptions about HIV. ____% of the women reported knowing at least one person in the community with HIV, and ____% stated that a member of the household has been diagnosed with HIV. ____% self-reported an HIV diagnosis, though HIV tests were not administered to confirm this value. ___% correctly identified that HIV is sexually transmitted. Within one week after the interview, the 50 women in each village gathered for a 2-hour lecture educational intervention that was implemented by local community health workers who are employed by a local primary care clinic. After the educational intervention, ____% correctly identified that HIV is sexually transmitted, and ____% demonstrated an understanding that condoms can help prevent HIV. ___% properly explained how to utilize a condom. The data demonstrated a statistically significant improvement in knowledge about HIV, though this study did not assess the application of this knowledge by the women participants. These results suggest that lecture-oriented educational programs implemented by community health workers may raise awareness about HIV prevention.

Why is this abstract considered high quality?

Every sentence in this abstract is supported by data.

• The first sentence now cites World Health Organization data to support the HIV prevalence statement.

• The research methodology is elucidated, including a clear statement that the women were self-selected, which may introduce bias into the study.

• Data is the focus of the abstract.

• Post-intervention data relates to specific educational points (e.g. how to use a condom) rather than whether the intervention was “beneficial.”

• The conclusion is supported by the abstract's data, and it is not overly broad. Based on the data obtained from the research, for example, one cannot state that "educating women is an effective way of increasing awareness about and reducing the prevalence of HIV." The research did not assess prevalence of HIV, nor can one assume that any form of education is effective at increasing awareness about HIV. Based on the data provided, however, one can state that the results suggest that the educational intervention described in the abstract may raise awareness about HIV prevention.

What is the difference between outputs and outcomes? Why must outcomes be included in descriptions of program models?

An example of an output is the number of program participants who are enrolled in a job training program. An example of an outcome is the number of program participants who have been able to coordinate a job due to the job training program. While outputs are important to track, evaluation needs to focus on measuring outcomes that reveal the extent and kinds of impact the project has on its participants. Impact could be reported in the amount of change in behavior, attitude, skills, knowledge, or condition of the target population. For example, an output would be the number of microfinance loans distributed or the number of loans repaid, whereas the outcome would be the number of microfinance participants who have significantly increased their income or risen out of poverty due to the loans.  Additional details are described in depth at http://www.uniteforsight.org/global-health-university/global-health-metrics

Qualitative Research

Suggestions and instructions for submitting abstracts for qualitative research may be found here: http://www.uniteforsight.org/global-health-university/quantify-research