Concept: Capture-Recapture Method of Estimating Population Size

Introduction

This concept describes the capture-recapture method of estimating population size, as it was used to estimate the prevalence of chronic kidney disease (CKD) in Chartier et al. (2015). This includes a general description of the capture-recapture methodology and how this methodology was applied using administrative and laboratory data available from the MCHP Data Repository. Links to the technical details and key findings from this research are also provided.

The capture-recapture method is used for "... estimating the size of a target population or a subset of this population that uses overlapping and presumably incomplete but intersecting sets of data about that population. Though the capture-recapture methods have some limitations, they are useful to estimate numbers of cases and numbers at risk in elusive populations". (Last, 2001).

Methodology

In Chartier et al. (2015) they used the capture-recapture method as one of three methods to estimate the prevalence of chronic kidney disease (CKD) in both adults and children, based on cases defined in administrative and laboratory data. Please read the Chronic Kidney Disease (CKD) concept for information on how CKD was defined using administrative and laboratory data.

The Chapman Estimator (Chapman, 1951)

In Chartier et al. (2015) they used the Chapman estimator to estimate CKD prevalence. This approach is often used in estimating the size of animal populations. This capture-recapture method was applied to the CKD data using the following general formula:




where:

N = estimate of the population size
n ₁ = size of sample 1
n ₂ = size of sample 2
m ₂ = number in sample 2 that are also in sample 1

For more information on this method, please read the section titled Chronic Kidney Disease Prevalence in the report.

MCHP Adaptation of the Chapman Estimator

Using the administrative data definition and lab-based definition, Chartier et al. (2015) estimated the total CKD population using the following formula:




where:

N _CKD = estimate of total CKD population
Admin _CKD = number meeting the administrative definition April 1 2009 - March 31 2012
Lab _CKD = number meeting the lab-based definition April 1 2006 - March 31 2012
Both _CKD = number meeting both definitions

NOTE: The administrative data used to identify CKD included the Hospital Abstracts, Medical Services (Physician Claims) and Drug Program Information Network (DPIN) data. For more information on how CKD was defined, please read the Chronic Kidney Disease (CKD) concept.

For more information on how the capture-recapture method was applied in this research, please read the capture-recapture technical definition and formula from the Appendix in the report.

Findings and Results

The following results are available from the report:


• Figure 3.2: Prevalence of Adults with Chronic Kidney Disease by Region and Estimation Method
  
• Table 3.2: Prevalence of Chronic Kidney Disease in Children by Estimation Method

For more general information, please read Chapter 3: Cohort Description: Chronic Kidney Disease and End Stage Kidney Disease which includes the key findings regarding the adult and child cohorts from this report, and specific details about these two groups.