High-level monthly, quarterly and annual turn-around time (TAT) reports are used to assess laboratory performance across the National Health Laboratory Service in South Africa. Individual laboratory performances are masked by aggregate TAT reporting across network of testing facilities.
This study investigated weekly TAT reporting to identify laboratory inefficiencies for intervention.
CD4 TAT data were extracted for 46 laboratories from the corporate data warehouse for the 2016/2017 financial period. The total TAT median, 75th percentile and percentage of samples meeting organisational TAT cut-off (90% within 40 hours) were calculated. Total TAT was reported at national, provincial and laboratory levels. Provincial TAT performance was classified as markedly or moderately poor, satisfactory and good based on the percentage of samples that met the cut-off. The pre-analytical, testing and result review TAT component times were calculated.
Median annual TAT was 18.8 h, 75th percentile was 25 h and percentage within cut-off was 92% (
Masked TAT under-performances were revealed by weekly TAT analyses, identifying poorly performing laboratories needing immediate intervention; TAT component analyses identified specific areas for improvement.
In South Africa, public health facilities across 52 districts provide patient care through primary healthcare (PHC) services, district, regional and tertiary hospitals. A wide spectrum of tests can be requested and submitted to the nearest pathology laboratory of the National Health Laboratory Service (NHLS). The NHLS is the choice laboratory service provider of the South Africa National Department of Health (NDoH). A network of more than 266 laboratories are strategically placed around the country to optimally accommodate the needs of local communities (urban and rural).
HIV-associated tests like HIV viral load (VL) and CD4 counts, like all NHLS laboratory tests, have strict predetermined organisational TAT cut-offs, set to reflect treatment guidelines requirements and standards of care for HIV management by local authorities and the World Health Organization (WHO)
The accurate reporting of TAT depends on the quality of data collected through the LIS, that is, the inclusion of automated system date and timestamps at various time points in the journey from patient venesection to result review.
Laboratory test TAT in the NHLS is monitored at national, provincial and laboratory level, with annual,
The aim of this article is to describe how weekly review of CD4 TAT analysis can enable the identification of non-compliant laboratories to facilitate effective and timely corrective action and ensure continuous quality management for improved service delivery. Data analysed represents performance prior to the national implementation of the weekly TAT dashboard.
Ethics clearance was obtained from the University of the Witwatersrand (M1706108). No patient identifiers were used for this study and laboratories and provinces were anonymised.
CD4 TAT data were extracted from the corporate data warehouse for the financial period April 2016 to March 2017 (2016/2017 financial year) for 46 CD4 testing laboratories. Total TAT was calculated for each sample tested and reported for 52 weeks, together with the TAT component data.
Data analysis included the calculation of the median, 75th percentile and the percentage of samples with a TAT within the stipulated organisation cut-off per week. This was reported per laboratory and per province (aggregated data of laboratories within each of the nine provinces). Performance classification was introduced at provincial level and based on the percentage samples within TAT cut-off as follows: (A) ≥ 95%: good performance, (B) 90.0% – 94.9%: satisfactory performance; (C) 85.0% – 89.9%: moderate to poor performance and (D) 80.0% – 84.99%: poor performance. Performance thus refers to the degree of compliance with NHLS TAT cut-off. The number of weeks that provinces and laboratories did not achieve the 40 h cut-off was reported. Outlier weeks were defined as weeks where the total TAT of all samples tested did not achieve 90% with a TAT under 40 h. Additional data analysis was done on the weekly laboratory data to describe the TAT component contribution to total TAT per laboratory per week and included: (1) lab-to-lab TAT, (2) reg-to-test TAT and (3) test-to-review TAT. The target times set for each TAT component are (1) 14 h, (2) 24 h and (3) 2 h. Although TAT component analysis by laboratory is distributed weekly, for this study, only specific laboratories were selected to represent different levels of compliance and performance categories to demonstrate how individual TAT components affects total TAT. Outlying laboratory TAT components (> 24 h and < 2 h) were correlated with Beckman Coulter engineer logs to verify the impact of instrument downtime on prolonged TAT (data not shown).
Laboratory site visits were conducted to assess root cause analysis for below standard TAT (< 90% processed for > 40 h) performance identified.
Data were prepared and analysed using SAS version 9.4 (Cary, North Carolina, United States) and GraphPad software (San Diego, California, United States). The nine provinces were numbered 1–9, with individual laboratories within a province assigned a number and labelled accordingly (i.e. 1.5 represents province 1 and laboratory 5). Box and whisker plots were created for individual laboratory data over 52 weeks. National total test volumes and TAT was plotted against the 50th and 75th percentiles in a bar graph. Provincial total TAT was plotted as 75th percentile per performance category per week. Individual laboratory distribution of 75th percentiles per performance category was plotted, indicating high (> 350 samples per day), medium (150-350 samples per day) and low volume facilities (< 150 samples per day). Component TAT was plotted as stacked bar graphs, showing the 75th percentile for pre-analytical, testing, and review TAT for selected laboratories representing the four performance categories.
In this study 3 332 599 CD4 test TAT were analysed. For the 2016/2017 financial year, the national median TAT for all CD4 tests was 18 h with a 75th percentile of 23 h (
National annual National Health Laboratory Service CD4 total turn-around time for all samples tested during the 2016/2017 financial year.
Laboratory TAT component | Total TAT |
Target TAT (hours) | 75th percentile minimum to maximum | |
---|---|---|---|---|
Median | 75th percentile | |||
Total TAT | 18.8 | 23.0 | 40 | 10–49 |
Laboratory-to-laboratory | 6.3 | 10.0 | 14 | 0–24 |
Registration-to-test | 17.3 | 22.0 | 24 | 9–43 |
Test-to-review | 1.3 | 1.6 | 2 | 0–8 |
TAT, turn-around time; CD4, cluster of differentiation 4.
The weekly distribution of the 50th percentile (median) and 75th percentile showed good consistency despite fluctuations in test volumes across the network of testing laboratories (
National total turn-around time of National Health Laboratory Service CD4 tests per week for the 2016/2017 financial year. The 50th percentile (median, green circles), 75th percentile (red circles) and volume of samples (grey bars) are depicted.
Annual global TAT distribution did not identify any poor performance over 52 weeks. To identify poor performances, national TAT were analysed per province. The number of CD4 tests ranged from 65 395 (lowest) to 1 066 137 (highest) (
Annual provincial National Health Laboratory Service CD4 data, indicating test volumes, the 75th percentile total turn-around time, the percentage of samples within turn-around time cut-off and the number of representative laboratories for 2016/2017 financial year.
Province | Number of CD4 tests (52 weeks) | Total TAT |
% samples within 40 hour TAT cut-off | Classification based on % samples within TAT cut-off (Group) | Number of CD4 laboratories per province | |
---|---|---|---|---|---|---|
75th percentile | Min to max range (hours) | |||||
1 | 352 277 | 23.02 | 18–37 | 92 | 90.0–94.9 (B) | 6 |
2 | 181 391 | 25.14 | 16–35 | 90 | 90.0–94.9 (B) | 3 |
3 | 756 661 | 21.75 | 20–27 | 95 | ≥ 95 (A) | 6 |
4 | 1 066 137 | 23.38 | 21–28 | 93 | 90.0–94.9 (B) | 11 |
5 | 235 772 | 34.18 | 18–80 | 82 | 80.0–84.9 (D) | 4 |
6 | 280 985 | 28.80 | 20–43 | 87 | 85.0–89.9 (C) | 5 |
7 | 65 395 | 17.58 | 13–23 | 98 | ≥ 95 (A) | 3 |
8 | 172 295 | 18.85 | 14–40 | 94 | 90.0–94.9 (B) | 4 |
9 | 221 686 | 20.87 | 17–27 | 95 | ≥ 95 (A) | 4 |
Performance classification: A: good performance; B: satisfactory performance; C: moderate to poor performance; D: markedly poor performance.
TAT, turn-around time; CD4, cluster of differentiation 4; Min, minimum; max, maximum.
Three provinces (3, 7 and 9) were classified as category A (good performance). These laboratories were able to maintain all CD4 reporting within organisation-stipulated TAT at greater than 95% and 75th percentile TAT of 21.7 h (province 3), 17.6 h (province 7) and 2.08 h (province 9). Four provinces (1, 2, 4 and 8) were categorised B (satisfactory performance) having 90% – 94.9% of samples meeting the TAT cut-off; 75th percentile values for these provinces were 23 h, 25.1 h, 23.4 h and 18.9 h, respectively. Two provinces failed to meet the TAT cut-off and were classified as categories C (moderate poor performance; province 6) and D (markedly poor performance province 5), indicating that less than 90% of samples met the CD4 TAT. Within the latter provincial performance clusters (C and D), the 75th percentile reported was 28.8 h and 34.2 h.
No weekly outliers (weeks where total TAT did not meet 90% < 40 h) were noted in the three good performance provinces (3, 7, and 9; comprising
Weekly national 75th percentile CD4 turn-around time of nine provinces per performance category for the 2016/2017 financial year. (a) good performance (
The different performance levels identified at provincial level still masked the performance and contribution of individual laboratories to provincial performance. Scatter plots were constructed to visualise the performance of individual laboratories over 52 weeks per provincial performance category. Results showed that irrespective of the provincial performance classification (
Scatter plots of the 75th percentile total turnaround time of individual laboratories in the National Health Laboratory Service within the provincial performance classification groups A to D for the 2016/2017 financial year. The overall 75th percentile total turn-around time per laboratory is indicated above each plot. (a) good performance (
The 75th percentile across 52 weeks for the good performance provinces (3 provinces and 13 laboratories) showed good overall compliance (tight clumping of weekly 75th percentile values) where the overall 75th percentile for the whole period ranged from 5.9 h (laboratory 7.1) to 24.8 h (laboratory 9.2) (
The number of weeks that laboratories did not meet the cut-off criteria of 90% with TAT under 40 h varied among categories and laboratories (0–22 weeks), where 12 of 46 laboratories (irrespective of performance category) had zero outlying weeks (26%), 25/46 (54%) more than 5 outlying weeks and 6 (13%) between 6 and 10 outlying weeks. Only three laboratories showed outliers for more than 10 weeks where cut-off was not met (6.5%).
Stacked bar graphs showing examples of individual laboratory weekly performance for 2016/2017 financial year. 75th percentile turn-around time components color-coded: lab-to-lab (orange), reg-to-test (blue) and test-to-review (green), with cut-off of 40 h (red dotted lines). (a) good performance (laboratory A); (b) satisfactory performance (laboratory B); (c) moderate to poor performance (laboratory C) and (d) markedly poor performance (laboratory D).
The laboratory represented in
The laboratory contributing the most outlying weeks to group D (
TAT remains a key performance indicator of laboratory service efficiency.
Ideally, sample-by-sample real-time reporting of TAT would be the preferred way to monitor and assist laboratories in the identification of specific service delivery and related TAT challenges. Hierarchical global overview (usually annual) TAT reporting is however the simplest and most widely used, but masks poor performance, as confirmed by data from this study. Interrogating the weekly data by drilling down to laboratory level at weekly intervals, enables the identification of outliers and poor performers. This study showed that lower hierarchical levels, as well as shorter time periods, can unveil problematic and inefficient testing laboratories (
Summary of study findings.
National total TAT and TAT component results paint a picture of good overall performance for a specific test (i.e. CD4), masking underlying performance outliers.
Analysing smaller sections of data (province and laboratory) for shorter test periods (month and week) reveals the true performance level based on compliance with organisational test cut-offs.
Performance can be classified as good, satisfactory or poor based on the percentage of samples within the test cut-off and the mean 75th percentile total TAT reported.
Weekly TAT component analysis at laboratory level reveals possible causes of delayed total TAT and provides some insight into the type of intervention required.
TAT, turn-around time; CD4, cluster of differentiation 4.
The most common issues that impacted on prolonged lab-to-lab (pre-analytical) TAT were delays in transport or sample collection from clinics to testing laboratories and changes in courier routes and pick up times. Reg-to-test times (laboratory TAT) were prolonged due to instrument downtime, lack of trained staff to operate testing instruments, delayed sample registration in the testing facility, challenges with reagent availability and timely delivery, and staff strikes.
Data reported here demonstrate the need for more frequent TAT reporting in effective performance management.
Based on the data presented in this study, further refinement of the current reporting platforms is recommended to include daily reporting for rapid proactive intervention. A further recommendation is to integrate daily quality control tests, external quality assessment testing and equipment downtime supplier call-out data into the reporting to assist with focused troubleshooting and interventions.
Turn-around time monitoring and reporting are however guided by the requirements of the end user and will continue to be available at various time intervals for laboratory network management to assess overall trends, with weekly or daily reports to laboratory and programme managers enabling timely proactive intervention to ensure optimal laboratory performance and timely patient result reporting.
The monitoring of TAT in the NHLS is currently limited as end-to-end sample tracking system is absent. The TAT reported in this article thus only represents the time from first registration on the LIS to review of the result.
National and provincial analysis of TAT mask individual laboratory performance; therefore, TAT analysis by week and by laboratory is recommended to highlight laboratory TAT inefficiencies. Root-cause analyses were able to identify pre-analytical, analytical or post-analytical factors contributing to performance. TAT data was used to categorise performance at the provincial and laboratory level.
This study used the concept of zooming in to lower levels and shorter times of TAT reporting to identify possible non-compliant laboratories. In conclusion: (1) outlying weeks are not prescribed by provincial or laboratory classification of performance, (2) performance did not correlate to the size of the laboratory (i.e. test volumes of high, medium and low), (3) there were laboratories that had no outlier weeks during the analysed period that can serve as model laboratories for setting performance standards and good reproducibility of week-on-week performance across a network of testing laboratories.
This article was in part presented as a poster at the African Society for Laboratory Medicine meeting in 2018, Abuja, Nigeria (ID:PS-2-3b-070), 10–13 December 2018. The authors thank area, business and laboratory managers in the NHLS for their feedback on the use of the weekly TAT dashboard. Thank you to the Central Data Warehouse for the availability of data.
The authors have declared that no competing interests exist.
D.K.G. supervised the study by providing leadership and oversight as the project leader. L.-M.C. and N.C. designed the study, developed the methodology and conducted the research, data analysis, initial write-up and review. D.K.G. reviewed the data, provided editorial comments and technical input. All authors contributed to the manuscript development.
This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.
Data is available as online Supplementary Document 1.
The views and opinions expressed in this article are those of the authors and do not necessarily reflect the official policy or position of any affiliated agency of the authors.