The Structured Abstract

1. The Structured Abstract Session 5 C507 Scientific Writing

2. Functions of an Abstract • Provide the reader with information regarding the research performed • This may determine if the paper is read • Serves as a “miniature” of the research

3. Prose Abstract • Until the last few years, and still in some cases, abstracts were prepared as a short prose summary of the article • But, there was trouble in paradise…

4. The Problem • Squires BP. Abstracts: the need for improvement. Can Med Assoc J 1991;144:421 • 56% of abstracts did not report the technical design of the study • 79% did not describe how subjects were selected • 86% did not address limitations • 93% made no recommendations for future study

5. The Answer • The purpose of a structured abstract is to ensure that all the appropriate information is included in the abstract so that readers are provided full details of the work, which can save cost and time in their own work.

6. A Bit of History • Initial impetus came from Brian Haynes of the Ad Hoc Working Group for Critical Appraisal of the Medical Literature • Ad Hoc Working Group for Critical Appraisal of the Medical Literature. A proposal for more informative abstracts of clinical articles. Ann Intern Med 1987;106:598-604

7. And a Bit More • In 1990, Haynes and colleagues assessed the first few years of the use of structured abstracts • Reporting errors were decreased but not absent • Support from the scientific community was widespread

8. Instructions • No more than 250 words • Use the following general headings: • Objective • Design • Setting • Patients • Interventions

9. Instructions (Cont.) • Main Outcome Measures • Main Results • Conclusions

10. Objective • The abstract should begin with a clear statement of the precise objective or question addressed in the report. If more than one objective is addressed, the main objective should be indicated and only key secondary objectives stated. If a priori hypothesis was tested, it should be stated.

11. Design • The basic design of the study should be described. The duration of follow-up, if any, should be stated. • As many of the following terms as apply should be used:

12. For Intervention Studies • Randomized controlled trial • Non-randomized controlled trial • Double-Blind • Placebo Control • Crossover Trial • Before-After Trial

13. For Studies of Screening and Diagnostic Tests • Criterion standard (that is, a widely accepted standard with which a new or alternative test is being compared; this is preferred to the term “gold standard.”) • Blinded comparison • Masked comparison

14. For Studies of Prognosis • Inception cohort (subjects assembled at a similar and early time in the course of the disorder and followed thereafter) • Cohort (subject followed forward in time, but not necessarily from a common starting point) • Validation cohort or sample (if study involves modeling of clinical predictions)

15. For Studies of Causation • Randomized Controlled Trial • Cohort • Case-Control • Survey (not “cross-sectional study.”)

16. For Descriptions of the Clinical Features of Medical Disorders • Survey • Case Series

17. For Studies that Include a Formal Economic Evaluation • Cost-Effectiveness Analysis • Cost-Utility Analysis • Cost-Benefit Analysis • For new analyses of existing data sets, the data set should be named and the basic study design disclosed

18. Setting • To assist readers to determine the applicability of the report to their own clinical circumstances, the study setting should be described. • Of particular importance is whether the setting is the general community, a primary care or referral center, private or institutionalized practice or ambulatory or hospitalized care.

19. Patients or Other Participants • The clinical disorders, important eligibility criteria and key sociodemographic features of patients should be stated. • The number of participants and how they were selected should be provided, including the number of otherwise eligible subjects who were approached but refused.

20. Patients or Other Participants • If matching is used for comparison groups, characteristics that are matched should be specified. • In follow-up studies, the proportion of participants who completed the study must be indicated. • In intervention studies, the number of patients withdrawn for adverse effect should be given.

21. Patients or Other Participants • For selection procedures, these terms should be used” • Random sample • Population-based sample • Referred sample • Consecutive Sample • Volunteer Sample • Convenience sample

22. Intervention(s) • The essential feature of any intervention should be described, including their method and duration of administration. • The intervention should be named for its most common clinical name (usually this is so for drugs, but can also be true for specific chiropractic techniques).

23. Main Intervention Measures • The primary outcome measurements should be indicated as planned before data collection began. • If the paper does not emphasize the main planned outcomes of a study, this fact should be stated and the reason indicated

24. Main Intervention Measures • If the hypothesis being reported was formulated during or after data collection, this information should be clearly stated.

25. Results • The main results of the study should be given. • Measurements that require explanation for the expected audience of the article should be defined. • Important measurements not included in the presentation of results should be declared.

26. Results • It should be indicated whether observers were blinded to patient groupings, particularly for subjective measurements. • Where possible, results should be accompanied by confidence intervals, (ie, 95%) and the exact level of statistical significance.

27. Results • When risk changes or effect sizes are given, absolute values should be indicated so that the reader can determine the absolute as well relative impact of the finding.

28. Results • Where appropriate, studies of screening and diagnostic tests should use the terms sensitivity, specificity and likelihood ratio • If predictive values or accuracy are given, prevalence or pretest likelihood should be given as well.

29. Results • No data should be reported in the abstract that do not appear in the rest of the article.

30. Conclusion • Only those conclusions of the study that are directly supported by the evidence reported should be given, along with their clinical application (avoiding speculation and over-generalization) and an indication whether additional study is required before the information should be used in usual clinical settings.

31. A Few General Comments • For brevity, you do not need to use full sentences in all cases and for all entries. • This allows for more information to be provided per unit of space.

32. Structured Abstract for Literature Reviews • Normal organization is: • Objective • Data Source • Study Selection • Data Extraction • Data Synthesis • Conclusions

33. Objective • The focus of this statement should be guided by whether the review emphasizes factors such as cause, diagnosis, prognosis, therapy or prevention, • It should include information about the specific population, intervention or exposure and test or outcome being reviewed.

34. Data Sources • Provide a summary of the data sources you used, along with any time restrictions. • When a data base is used, the exact indexing terms used for article retrieval should be stated, including constraints (ie, English language only).

35. Study Selection • Describe the criteria used to select studies for detailed review from among studies identified as relevant to the topic. • Details should include particular populations, interventions, outcomes or methodological designs.

36. Data Extraction • Guidelines used for abstracting data and assessing data quality and validity should be described. • The method by which the guidelines were applied should be stated, for example, independent extraction by multiple observers.

37. Data Synthesis • The main results of the review, whether qualitative or quantitative, should be stated. • Methods used to obtain these results should be outlined.

38. Data Synthesis • Meta-analyses should state the major outcomes that were pooled and include odds ratios or effect sizes, and if possible, sensitivity analyses. • Numerical results should be accompanied by CIs, if applicable, and exact levels of statistical significance.

39. Data Synthesis • Evaluations of screening and diagnostic tests should address issues of sensitivity, specificity, likelihood rations, receiver operating characteristic curves and predictive values. • Assessments of prognosis could include summaries of survival characteristics and related variables.

40. Data Synthesis • Major identified sources of variation between studies should be stated, including, for example, differences in treatment protocols, cointerventions, confounders, outcome measures, length of follow-up and drop-out rates.

41. Conclusion • The conclusions and their applications should be clearly stated, limiting generalizations to the domain of the review. • The need for new studies may be stated.

42. Abstract for a Case Report • General headings include: • Objective • Clinical Features • Intervention and Outcome • Conclusion

43. Objective • Clearly state what you expect the reader to take from reading your paper • Could be to describe an interesting case, an unusual diagnosis, a case with an interesting response to therapy, etc.

44. Clinical Features • List only those clinical features that are specific to the case; do not provide a list of normal findings • Provide the reason why the patient sought care • Where possible, provide a diagnosis

45. Intervention and Outcome • Provide detail as to the treatment offered. • Detail should be sufficient for the reader to understand what was done • Provide detail as to the patient response; list outcome measures used and changes that occurred

46. Intervention and Outcome • Where the purpose of the case does not deal directly with a response to care (ie. a case that reports an unusual radiographic finding) provide information about referral, where appropriate.

47. Conclusion • Case reports cannot be generalized beyond the results from the single case; be cautious not to discuss effectiveness, efficiency or make broad generalizations • Be sure that your conclusion is supported by evidence from the case