Results of continuous monitoring of the performance of rubella virus IgG and hepatitis B virus surface antibody assays using trueness controls in a multicenter trial

A modified quality control protocol for infectious disease serology based on the Westgard rules

When traditional statistical quality control protocols, represented by the Westgard protocol were applied to infectious disease serology, the rejection limits were questioned because of the high rejection probability. We first define the probability of false rejection (Pfr) and error detection (Ped) for infectious disease serology. QC data in 6 months were collected and the Pfr of each rule in the Westgard protocol and Rilibak protocol was evaluated. Then, as improvements, we chose different rules for negative and positive QC data to constitute an asymmetric protocol, furthermore, while reagent lot changes, the mean value of QC protocol is reset with the first 15 QC results of new lot reagent. QC materials and Standard Reference Materials were tested synchronously in the next 6 months, to verify whether the Pfr and Ped of the asymmetric protocol could meet the requirement. Protocol 1 exhibited the higher level of rejection rate among the two protocols, especially after reagent lot changes; Pfr below the lower control limit (LCL) was 1.39-21.78 times higher than the upper control limit (UCL); false rejections were more likely to occur in negative QC data, with Pfr-total of 27-65%. The asymmetric protocol can significantly reduce the proportion of analytes with Pfr by over 20%. Systematic error due to reagent lot changes and random error due to routine QC data variation were considered potential factors for excessive Pfr. Asymmetric QC protocol that can reduce Pfr by different control limits for negative and positive QC data.

Does a change in quality control results influence the sensitivity of an anti-HCV test?

Background

Laboratories use quality control (QC) testing to monitor the extent of normal variation. Assay lot number changes contribute the greatest amount of variation in infectious disease serology testing. An unexpected change in six lots of an anti-HCV assay allowed the determination of the effect these lot changes made to the assay’s clinical sensitivity.

Methods

Two sets of seroconversion samples comprising of 44 individual samples and 9 external quality assessment scheme (EQAS) samples, all positive to anti-HCV, were tested in affected and unaffected assay lots, and the difference in the quantitative and qualitative results of the samples was analyzed.

Results

Of 44 low-positive seroconversion samples tested in affected and unaffected assay lots, only three samples had results reported below the assay cutoff when tested on two of the six affected assay lot. A further sample had results below the cutoff for only one affected lot. None of the EQAS samples reported false-negative results. Samples having a signal to cutoff value of less than 6.0 generally had lower results in the affected lots compared with the unaffected lots.

Conclusions

Unexpected changes in QC reactivity related to variation, in particular assay lot changes, may affect patient results. This study demonstrated that QConnect Limits facilitated the detection of an unexpectedly large variation in QC test results, allowed for the identification of the root cause of the change, and showed that the risk associated with the change was low but credible. The use of evidence-based QC program is essential to detect changes in test systems.

Comparison of four methods of establishing control limits for monitoring quality controls in infectious disease serology testing

BACKGROUND

A general trend towards conducting infectious disease serology testing in centralized laboratories means that quality control (QC) principles used for clinical chemistry testing are applied to infectious disease testing. However, no systematic assessment of methods used to establish QC limits has been applied to infectious disease serology testing.

METHODS

A total of 103 QC data sets, obtained from six different infectious disease serology analytes, were parsed through standard methods for establishing statistical control limits, including guidelines from Public Health England, USA Clinical and Laboratory Standards Institute (CLSI), German Richtlinien der Bundesärztekammer (RiliBÄK) and Australian QConnect. The percentage of QC results failing each method was compared.

RESULTS

The percentage of data sets having more than 20% of QC results failing Westgard rules when the first 20 results were used to calculate the mean±2 standard deviation (SD) ranged from 3 (2.9%) for R4S to 66 (64.1%) for 10X rule, whereas the percentage ranged from 0 (0%) for R4S to 32 (40.5%) for 10X when the first 100 results were used to calculate the mean±2 SD. By contrast, the percentage of data sets with >20% failing the RiliBÄK control limits was 25 (24.3%). Only two data sets (1.9%) had more than 20% of results outside the QConnect Limits.

CONCLUSIONS

The rate of failure of QCs using QConnect Limits was more applicable for monitoring infectious disease serology testing compared with UK Public Health, CLSI and RiliBÄK, as the alternatives to QConnect Limits reported an unacceptably high percentage of failures across the 103 data sets.

Development and Implementation of Autoverification Rules for ELISA Results of HBV Serological Markers

Autoverification is a process of using computer-based rules to verify clinical laboratory test results without manual review. But to date, there are few published articles on the use of autoverification over the course of years in a clinical laboratory. In our study, we firstly described the development and implementation of autoverification rules for enzyme-linked immunosorbent assay (ELISA) results of hepatitis B virus (HBV) serological markers in a clinical immunology laboratory. We designed the autoverification rules for HBV by using Boolean logic on five clinically used serological markers in accordance with the framework of AUTO-10A, issued by the American Clinical Laboratory Standards Institute in 2006. The rules were written into the laboratory information system (LIS) and installed in the computer, so we could use the LIS to screen the test results. If the results passed the autoverification rules, they could be sent to doctors immediately. To evaluate the autoverification rules, we applied the real-time data of 11,585 patients with the autoverification rules. The autoverification rate of the five HBV serological markers was 79.5%. Furthermore, the turnaround time (TAT) was reduced by 38% (78 minutes vs. 126 minutes). The error rate was nearly eliminated. These results show that using LIS with autoverification rules can shorten TAT, enhance efficiency, and reduce manual review errors.

Determination of quality control limits for serological infectious disease testing using historical data

BACKGROUND

An effective quality control (QC) program requires the establishment of control limits within which the results of the QC sample is expected to fall. Traditionally, the mean plus/minus two standard deviations calculated for a set of QC sample results is used to establish control limits. Allowable total error (TEa) and Westgard rules aid in interpreting QC sample results. Westgard rules assume QC sample results are normally distributed and TEa assumes commutability between the QC sample and patient results. None of these paradigms apply to infectious disease testing.

METHODS

RESULTS from the NRL's QC program were extracted and sorted into assay/QC lot number-specific data. Control limits for selected QC samples used to monitor 64 commonly used serological assays were calculated and validated using the within- and between-QC lot variance of data from each of the assay/QC combinations.

RESULTS

No assay/QC combination had more than 10% of results less than the lower control limit or greater than the upper control limit. Of the 423 assay/QC lot combinations, 14 (3.3%) had more than 5% of results less than the lower limit and 48 (11.3%) had more than 5% of results greater than the upper limit calculated for that assay/QC combination.

CONCLUSIONS

The control limits, established by this novel method, are based on more than a decade of QC test results from >300 laboratories from 30 countries and provides users of the NRL QC program evidence-based control limits that can be applied in isolation or in conjunction with more traditional methods for establishing control limits.

Determining rubella immunity in pregnant Alberta women 2009-2012

Rubella IgG levels for 157,763 pregnant women residing in Alberta between 2009 and 2012 were analyzed. As there have been no reported cases of indigenous rubella infection in Canada since 2005, there has been a lack of naturally acquired immunity, and the current prenatal population depends almost entirely on vaccine induced immunity for protection. Rubella antibody levels are significantly lower in younger maternal cohorts with 16.8% of those born prior to universal vaccination programs (1971-1980), and 33.8% of those born after (1981-1990) having IgG levels that are not considered protective (<15 IU/mL). Analysis across pregnancies showed only 35.0% of women responded with a 4-fold increase in antibody levels following post-natal vaccination. Additionally, 41.2% of women with antibody levels <15 IU/mL had previously received 2 doses of rubella containing vaccine. These discordant interpretations generate a great deal of confusion for laboratorians and physicians alike, and result in significant patient follow-up by Public Health teams. To assess the current antibody levels in the prenatal population, latent class modeling was employed to generate a two class fit model representing women with an antibody response to rubella, and women without an antibody response. The declining level of vaccine-induced antibodies in our population is disconcerting, and a combined approach from the laboratory and Public Health may be required to provide appropriate follow up for women who are truly susceptible to rubella infection.