Recommendations for the integration of standardized quality indicators for glucose point-of-care testing

Internal QC QI

1. Total number of QC points

2. Number of QC points outside acceptability limits

3. QC acceptability limits

4. Source of QC acceptability limits

5. Number of glucose meters

6. Number of test strip lots

7. Number of QC lots

Operator certification QI

1. Total number of operators

2. Total certified operators

3. Total operators with pending certification

4. Total expired operators

5. Total soon to expire operators

EQA QI

1. Number of EQA tests performed

2. Number of EQA flags

Field validation of the two other QIs, positive patient identification critical results follow-up was conducted in our previous manuscript [3].

PPID QI

This indicator has been discussed in detail previously [3] and is now included in the Canadian QI comparison program initiated by the Quebec Society of Clinical Biology [6] in collaboration with the Canadian Society of Clinical Chemists and the Working Group on Laboratory Errors and Patient Safety (WG-LEPS) of the IFCC. Figure 1 summarizes the findings from the past six QI data submission events from the Canadian program between February 2022 and January 2024.

Figure 1:

Average % of POC glucose tests performed without valid PPID across six data submission events. The number of laboratories included for each event is indicated in brackets next to the event date. Data are divided by sites that have admission, discharge and transfer system (ADT) connectivity for patient demographics with the glucose meters and sites that do not.

Operator certification QI

Data on operator certification status for glucose were submitted by 13 sites and are summarized in Table 3 below. Not all sites were able to separate expired from pending certifications and some sites were not able to identify operators that were soon to expire. The total operators are defined as the operators that are currently certified and including operators with pending certification and soon to expire. The percentage of certified operators was calculated for each site, ranging from 50–100 %. The 25th percentile was calculated as 81 % and the 75th percentile as 99 %.

Internal QC QI

Preliminary data for this QI were obtained from the working group members over a one-month period. Data were submitted from 22 hospital sites across Canada, including sites using both the Nova StatStrip (n=12) and Roche Inform II (n=10) glucose meters. The data are shown in Table 3. As per the IFCC WG-LEPS recommended process, the 25th and 75th percentiles were calculated for each meter type based on the submitted data [7]. For the Nova StatStrip meter, the 25th and 75th percentiles of performance were 0.40 % and 0.57 %, respectively. For the Roche Inform II, the 25th and 75th percentiles were calculated as 0.38 % and 0.81 %, respectively. Most sites indicated using the QC limits provided by the manufacturer with one site using limits calculated from in-house performance of the glucose meters (N4). The number of test strip lots and QC lots included in the analysis is indicated in Table 4 for each site. One month of QC data was extracted from POCT data management software by glucose meter for one submitting site as an example for further investigation, with the data summarized in Table 5 as an example report. Obvious QC vial switches are noted for meters 1 and 7. QC were performed at least 31 times on each meter for the month, indicating QC performed at least every 24 h as per the provided site operating procedure.

EQA QI

EQA data were submitted from 20 sites and are summarized in Table 6 below. The percentage of EQA failures was calculated for each site. Given the low number of failures, the 25th and 75th percentiles for performance could not be calculated.

PPID QI

The PPID QI was discussed in detail previously [3] This QI has been successfully integrated into the Canadian QI comparison program, with an average of 58 sites across Canada submitting data for each event. We have previously identified that sites with admission, discharge transfer (ADT) system connectivity for POCT glucose meters had less PPID errors than sites without ADT connectivity. This trend has been consistent across six QI submission events with the Canadian QI comparison program. This demonstrates the importance of connectivity for POCT devices to improve quality of results documentation.

Operator training QI

For devices that have connectivity to a POCT data management software, certification reports can be generated on a routine basis to calculate the QI using the number of operators with different certification status, including certified, expired and pending. Based on the data submitted for this QI, most sites show a high number of certified operators for POC glucose testing. A couple of sites had a relatively high number of operators with pending certifications. This may be related to different certification processes used by sites. For example, some sites may have one certification period for all operators whereas other sites may have rolling certifications. For devices that have operator lock-out capabilities, non-certified users will not be able to perform testing. It also highlights the risk of testing not being possible when required in an urgent situation, as when users’ certifications have expired or are pending. For manual POCT or devices that do not have operator lock-out capabilities, including manual POCT, monitoring of this QI may be important to understand the risk of testing by non-certified individuals. QI data may also prompt discussions around minimum recertification criteria and auto-certification based on testing frequencies. Many expired certifications may indicate high staff turnover rates and may suggest the need for alternative training opportunities to better support the program.

Internal QC QI

The rate of QC failure was higher in sites with Roche meters (75th percentile 0.81 %) compared to sites with Nova meters (75th percentile 0.57 %). This may be related to the number of QC points reported from Roche sites vs. Nova sites with twice as many points reported, on average, for Nova sites. This QI was chosen to ensure monitoring of the analytical testing process. QC testing is typically performed by clinical staff performing POC glucose testing, and QC failures can indicate errors associated with the testing process that may require follow-up with a clinical area or operator. Data from obvious QC vials switches was removed from the analysis. Vial switches represented 33–72 % of all QC failures across the sites. Aggregate data from all glucose meters in an institution may not be meaningful, particularly in institutions with large numbers of devices or variety of inpatient units as workflows specific to different specialties may impact successful QC differently. Analysis of data with more granularity (i.e. by glucose meter, unit, operator etc.) can identify more nuanced issues with the POCT device or program. Table 5 shows an example QC report by glucose meter for one site. Analysis by individual glucose meter allows for identification of specific meters with QC outside the acceptability limits, as well as instances with obvious QC vial switches. It can also recognize areas where infrequent QC is performed, which can prompt discussions with clinical areas about utilization.

EQA QI

EQA is a key component of laboratory quality assurance practices, including POCT. Similarly to EQA for central laboratory tests, EQA for POCT can help identify gaps in the testing processes. Data from submitting sites here showed a low rate of EQA failure. Where there were failures, only one site could explain the cause of the failure. Failures at other sites were deemed random. Ideally, clinical staff who perform patient testing are the ones also performing testing for EQA challenges. If laboratory staff observe clinical staff performing testing, this is an opportunity to identify errors with the process of testing technique, which could explain EQA failures. Otherwise, it can be difficult to identify a root cause. For programs with a very large number of devices (e.g. blood glucose monitoring) it may pose a challenge to enroll every device in an EQA program due to cost or logistical challenges. However, the advantage of including every device allows for complete intra-instrument comparisons.

Critical Results Follow-up QI. Preliminary findings for this QI were presented previously [3]. Briefly, this indicator monitors compliance with repeat of critical POCT glucose measurements within a defined period for confirmation. Based on experience in their respective sites, authors in the working group have indicated challenges with implementation of this QI. A recurring challenge is the fact that clinical staff and leadership do not agree with the laboratory questioning (critical) results without knowledge of the clinical context. However, clinical staff are not always cognizant of, or they do not always appreciate or recognize the impact of pre-analytical factors on results. The timing required for valid repeat of measurement has also been questioned. Sites planning to implement this QI are advised to provide data to clinical staff on the historical rate of discordant glucose results when critical results are repeated, if available. Based on data provided by the authors and published previously [3] the average rate of discordant repeat was 36 %. Studies in the literature have demonstrated that the majority of erroneously high POC glucose results are due to insufficient hand washing prior to testing or testing of specimens that are contaminated after being taken from a contaminated line [8].