Studies of method effectiveness must be carefully assessed for comparability of findings. Several parameters are identified that are important in the assurance of comparable results. This article discusses these issues with the use of data from previously published studies and emphasizes the management implications of use-effectiveness data.
PIP: The focus of the article on natural family planning (NFP) efficacy studies is on issues in data analysis and the use of efficacy data in management decisions. Data problems are identified as variable study designs, different times of entry, accuracy of self selection and self reporting, and consistency of method use and followup. Examples with actual data of these problems are given for Chile, Bangladesh, Kenya, and Korea. The methods used to measure effectiveness are not comparable. Clinical trials are prospective measures of efficacy, while population surveys are sometime as used for in-use efficacy. use efficacy may be calculated in 2 ways: the Pearl Index (the failure rate expressed as the number of pregnancies divided by the number of months of exposure and multiplied by 1200) and the cumulative life table method. Numerator problems occur due to reporting bias and loss of followup. Denominator problems suffer from self selection into the method, bias of recall, different times of entry into the study, and unreported concurrent use of other methods. Another variable is the median age of participants and the number of months of method use completed at the time date gathering began. The comparative example is given of data from Bangladesh, Kenya and Korea where entry into the study was late and related to the lowest unplanned pregnancy rate. The center with the youngest population also showed the lowest unplanned pregnancy rate. The timing is also related to whether the respondent is breastfeeding just after delivery or interbirth, where unplanned pregnancies are greater. It is pointed out that another method such as effectiveness assessment might be used. It measures the % of clients achieving adequate spacing or their stated goal. The advantages are in its being a personalized indicator, easier to assess, and provides excellent time series data. It is not useful in comparing methods or programs. Examples of usefulness of efficacy data are given; i.e., multiple decrement life tables can tell management about continuers versus discontinuers. The reason for the unplanned pregnancy may tell whether there was a misunderstanding. In the Chilean data set, most of the pregnancies occurred during lactation after the return of menses. When lactation is controlled for, the rates between spacers and limiters disappears. A common error is knowledgeable users do not follow the rules of the ovulation method and there is cultural resistance to revealing the number of sexual partners.