Handbook for Healthcare Auditors: Methods and Issues Related to Statistical Sampling

Healthcare auditors face unique issues. The audit populations and samples that they routinely deal with are not those discussed in survey sampling textbooks. In particular, their audit populations and samples generally contain a large number of zero values since audit populations tend to be mostly correct (and a zero “error” is recorded). The presence of a large number of zero errors is a very important feature of audit sampling and has many important consequences that affect the appropriateness of using many of the statistical techniques employed in survey sampling. This represents a unique feature that distinguishes audit sampling from most other types of surveys. Because most books on sampling theory do not treat this, Handbook for Healthcare Auditors: Methods and Issues Related to Statistical Sampling will cover this aspect of audit sampling in some detail and offer ways of coping in practice. This handbook provides an in-depth discussion of the various sampling designs available to healthcare (and other) auditors. The authors’ set of Excel macros is available to purchasers and is discussed throughout, including random number generation and sample projections using standard but also very-recently-developed methodology. Particular attention is paid to stratified designs (selecting the number of strata, determining strata boundaries, appraising the results) and multistage designs (single-stage, two-stage, three-stage). Read more