# Assessing laboratory specimen losses for the city of Johannesburg, South Africa

## TL;DR

This study assesses specimen losses in Johannesburg's pathology system using barcode data, finding most specimens are processed successfully but some are lost or rejected.

## Contribution

The study introduces a novel method for assessing specimen losses using barcode data from paper-based registers in a low-resource setting.

## Key findings

- 97.76% of barcodes had matching registered, tested, and reviewed dates.
- 5.79% of barcodes were rejected for one or more test codes.
- Sub-district D had the highest rejection rate at 10.93%.

## Abstract

Specimen losses across the pathology value chain (PVC) result in missed diagnostic opportunities. It is difficult to fully assess these due to the current paper-based systems, with tracking of specimens only possible on the laboratory information system (LIS).

This study aimed to assess specimen losses using the paper-based register.

Randomly selected Primary health care (PHC) facilities, City of Johannesburg, South Africa.

The retrospective descriptive study design was used to scan 1,000 barcodes from facilities in sub-districts A to G. Data was limited to barcodes from the request form and excluded surveillance testing. Matching data from the laboratory repository was extracted. PVC losses were assessed by determining the percentage of scanned barcodes that had a registered, tested, reviewed and/or rejected date. The analysis was stratified according to sub-district, health facility type and test code.

The dataset analysed included 33 867 barcodes with 121 697 test codes, equating to 3.59 tests per barcode. Matching registered, tested and reviewed dates were detected for 33 107 (97.76%) barcodes. In total, a rejection for one or more test codes was detected for 1,961 barcodes (5.79%). At the sub-district level, between 95.95% (D) and 98.90% (E) of barcodes were reviewed. The rejection rate ranged from 3.27% (F) to 10.93% (D). For community health centres and clinics, 97.37% and 97.97% of the barcodes had a matching reviewed date.

PVC losses reported were 4.05%, excluding rejections (5.79%), with slightly higher levels noted at the sub-district level.

The continuous audit of PVC losses is recommended.