|
 
 |
| ORIGINAL ARTICLE |
|
| Year : 2017 | Volume
: 8
| Issue : 1 | Page : 42-46 |
|
Establishing the reference change values (RCVs) and validating the delta check auto-verification in a clinical biochemistry laboratory
Denver Clive Fernandez1, SS Avinash1, M Malathi1, AR Shivashankara1, Arun Kumar1, Pearl Andrea Fernandez2
1 Department of Biochemistry, Father Muller Medical College, Mangaluru, Karnataka, India
2 St. Aloysius College Mangalore, Mangaluru, Karnataka, India
| Date of Web Publication | 2-Feb-2017 |
Correspondence Address:
Denver Clive Fernandez
Department of Biochemistry, Father Muller Medical College, Mangaluru, Karnataka
India
 Source of Support: None, Conflict of Interest: None  | Check |
DOI: 10.4103/0975-9727.199363
Aims: Establishing the reference change values (RCVs) and validating the delta check auto-verification in the hospital information system (HIS). Materials and Methods: This study was conducted in the Hospital Laboratory-Biochemistry. Fifty-one parameters were analyzed in three phases. Phase I: Delta check reference change values were established. Phase II: Delta check auto-verification was validated in the hospital information system. Phase III: Calculation of test requiring manual verification, true and false positive rates. Results: Out of all the test results, 1.35% failed the RCV-delta check thus requiring manual verification, and the remaining 98.65% were auto-verified. Only 0.12% test results were true positives indicating laboratory error, and 1.23% were false positives and were correlated clinically. Ten percent simulated data results and 0.37% actual patient results were not identified by the newly introduced HIS. Conclusions: RCV-delta check is a refined form of the delta checks used to analyze acceptable analytical and biological variation in laboratories. Majority of tests passed the RCV-delta check auto-verification, implying that very few test reports require manual verification. True positives can be detected in the laboratory. All HISs should be validated before implementing complete auto-verification. Keywords: Auto-verification, delta check, reference change values (RCVs), turnaround time (TAT)
How to cite this article:
Fernandez DC, Avinash S S, Malathi M, Shivashankara A R, Kumar A, Fernandez PA. Establishing the reference change values (RCVs) and validating the delta check auto-verification in a clinical biochemistry laboratory. Muller J Med Sci Res 2017;8:42-6 |
How to cite this URL:
Fernandez DC, Avinash S S, Malathi M, Shivashankara A R, Kumar A, Fernandez PA. Establishing the reference change values (RCVs) and validating the delta check auto-verification in a clinical biochemistry laboratory. Muller J Med Sci Res [serial online] 2017 [cited 2023 Jun 1];8:42-6. Available from: https://www.mjmsr.net/text.asp?2017/8/1/42/199363 |
| Introduction | |  |
A large clinical biochemistry laboratory catering to tertiary care patients is faced with challenges to dispatch accurate and precise patient test reports within an acceptable turnaround time (TAT). Reducing the TAT can be achieved by preanalytical, analytical, and postanalytical automation, and linking to laboratory information systems (LISs) and Hospital information systems (HISs).[1],[2],[3] Accuracy and precision of reports can be achieved by eternal quality control (EQC) and internal quality control (IQC), respectively. Further, checks of patient test reports in the form of limit check, critical value checks, delta checks, and consistency checks can significantly improve the quality of results.[4] Manual verification of all the above in medium and large-sized laboratories delays the process, and makes applying these checks for all reports impractical. Auto-verification of these patient report checks can minimize the delay.[4]
Delta check is an important part of a patient test report check that ensures the release of results with acceptable biological variation but it does not take into account the analytical variations in the lab. Hence, the establishment of reference change value (RCV), which takes into account the analytical as well as intra-individual variation, is of significance for any medical testing laboratory.[5] RCV/critical difference, is the value that must be exceeded before a change in consecutive test results, is statistically significant, at a predetermined probability.[5] RCV brings an objective evidence-based edge to the interpretation of a patient's test report when a previous report is available.
In a large clinical biochemistry lab with an input of more than 1,000 samples per day, it is absolutely necessary to establish the RCV. This present study was done in order to reduce the TAT for the release of accurate and precise results with very little unacceptable and unexplainable variation. The RCV and validating the delta check auto-verification manually according to the CLSI Auto-10A guidelines,[6] in a clinical biochemistry lab was done.
| Materials and Methods | | |
This study was conducted in the Clinical Biochemistry section of the Hospital Laboratory. Fifty-one parameters were analyzed.
Institutional ethical clearance was taken before starting the study. No human or animal subjects were directly involved in this study. This study was carried out in three phases.
Phase I
The delta check (RCV)[7] was established for each of the chemistry and immunoassay analytes and was done based on the following equation:
RCV = 2½ × Z × [CVA2 + CVI2]½
where RCV — Reference Change values [7]
2½ — Variation in present sample is compared with previous sample
Z — Z- score for 95% confidence interval.
CVA — from the IQC program of our laboratory
CVI — from Westgard database of desired biological variable [8]
Total variance in the lab,[5] SD 2 = preanalytical Standard Deviation (SD)2 + analytical SD 2 + postanalytical SD 2
Instead of SD, CV was used as a measure of dispersion.[5] Since the preanalytical, i.e. sample collection and centrifugation are standardized, the preanalytical variation is minimal and was excluded. Only Analytical CVA and CVI were used for the calculation of RCV according to Harris and Yasaka. In the IQC, all the analytes were estimated using IQC dispersed across 6 months. This IQC data was used for calculation of CVA. IQC controls that are freeze-dried lyophilized powder stored at –20° were reconstituted with 5 mL distilled water and restored in aliquots at –20° when required these aliquots were brought to room temperature by keeping out for 20 min and then the quality control (QC) run in the instrument, the similar procedure was followed for all analytes based on types of control used. CVA was calculated for all the parameters, using the following formula:  . The respective values were computed using the formula for RCV and the results obtained are shown in [Table 1].
Phase II
Validation of delta check auto-verification alert in the HIS was done under standard guidelines, CLSI Auto-10A guidelines,[6] which include:
- Using simulated data of patient test result, artificially generated data (30 values for each analyte) were entered in HIS and were verified manually.
- Actual patient values spanning over a month for each chemistry and immunoassay analyte were run in the system. The data generated were verified manually.
Validation of HIS with manual cross-checking was done to ensure the systems reliability.[4] Here the patient test values, simulated/actual values were fed in the test site for a patient, subsequent reports were generated with different values at a different time/date. Values that failed the RCV-delta check appeared with a yellow background in the HIS. The values that matched with manual verification were concluded as validated, andvalues that did not match with manual verification were concluded as validation failed. The results obtained are mentioned in [Table 2].
Phase III
Using the decision-making algorithm according to the CLSI Auto-10A guidelines and calculating the percentage of test requiring manual verification, true positive and false positive rates of the reporting were generated.
True positives are errors made in specimen identification, test performance, or test result reporting, which are correctable errors in the lab. False positives are changes ascribable to response in disease or therapy. The percentage of tests that were auto-verified by RCV-delta check alone and the percentage requiring manually verified were determined by checking with patient requisitions through the following algorithm based on the CLSI Auto10-A guidelines.[9]
| Results | | |
The RCV was established and the values obtained are mentioned in [Table 1].
Out of all the inpatient test results, 1.35% failed the RCV-delta check thus requiring manual verification; the remaining 98.65% were auto-verified [Table 2]. This implies that 1.35% patient test results fail the delta check when RCV are used, nevertheless, delta check is only a part of sample check and does not eliminate the need for limit, critical and consistency checks which need to be established and validated. Amongst the 1.35% delta fail 0.12% were true positives refer [Table 3]. True positives are errors identified in the laboratory, while the remaining 1.35% were correlated clinically, indicating that they were false positives. Moreover, 10% of actual patient result and 0.37% of simulated data result were not picked up by the newly introduced HIS in detecting delta errors.
| Discussion | | |
Most tertiary care laboratories rely on manual verification alone to validate test reports; this increases the TAT in medium- to large-sized laboratories and leads to inconsistent quality.
Delta check is an important part of sample check that ensures release of reports with acceptable variation. There are four types of delta check: Delta difference, delta percent change, rate difference, and rate percent change. The delta check has been considered a recognized form of quality assurance.[5] But it does not account for analytical variation in the laboratory, in present laboratory practice the analytical variation is readily available as coefficient of variation (CV) from the IQC program. RCV accounts for this analytical variation as CVA as well as the intra-individual variation (CVI), and is known to be a refined form of the delta check.[5] In this study, as a part of phase I, we established the RCV using the formula as mentioned above for 51 analytes done in our laboratory.
Using these established RCV for checking patient test report is a sign of good laboratory practice, but for most tertiary care centers this is not practical with the given sample load. Hence, it is necessary to implement auto-verification. Auto-verification maintains the quality of test reports and significantly reduces the TAT. Here we entered the RCV into the HIS, test site followed by the live site.
Auto-verification systems dispatch patient test reports after performing appropriate predefined checks without the need of manual validation. Therefore, before auto-verification is implemented, it needs to be validated.[6] The guideline [6] also recommends a two-step validation, with simulated data and actual patient test results. Simulated data help check this system across the entire analytical measuring range for robustness of performance and helps pick up errors not detectable by using patient test results alone. Validation with simulated data revealed 10% error in the HIS, whereas validation with patient test results revealed 0.37% errors, This discrepancy was seen because, validation with simulated data tests this system across it's analytical measuring range (AMR) thereby helps in detecting minute errors, these may not be seen in day-to-day practice but need to be identified and corrected. The errors identified were taken constructively by the HIS team and corrected.
Approximately, 1.35% of test results failed the delta check and required manual verification. Remaining 98.65% were auto-verified when the RCV-delta check was used alone, further checks of these patient reports in the form of limit, critical, and consistency checks are mandatory for complete auto-verification.[4],[6] Nevertheless, this clearly implies that very few test reports need manual validation, this is supported by a study done by Shih et al., who demonstrated that 80% test reports can be auto-released and the TAT is shortened.[4] Their study was conducted using a middleware and a dedicated auto-verification system. Their delta checks were based on the four types of delta checks as follows: Delta difference, delta percent change, rate difference, and rate percent change.[5] In our study, we used RCV as a form as delta check and observed it to be a form of delta percent change; our validation was performed using the HIS. We observed a higher auto-verification rate of 98.65% as we performed only the RCV-delta check. The purpose of delta check is to detect true positives. True positives are errors that occur in the laboratory [9] (for example, specimen mix-up, calibration errors, and typing errors). The true positive rate was 0.12%. False positives are changes occurring in patient test report either due to disease or due to response to treatment; these correlate clinically and is not the main purpose of the delta check. The common false positives encountered were cases of hypokalemia on treatment, nephrotic syndrome, Chronic Kidney Disease (CKD) patients on dialysis, and hepatic failure.
RCV-delta check was therefore a better model of the delta checks, which was established and validated in the HIS. All HISs need to be validated before the implementation. Auto-verification drastically reduces the TAT.
Standardization of patient test result verification with RCV-delta percentage change and auto-verification under the CLSI Auto-10A guidelines can produce quality reports consistently even as the test load increases with time.
Limitations
- One percent error was present in the HIS with actual patient results were tested.
- Ten percent error was present with simulated data.
- For establishing the RCV-delta check, we have used analytical CV. For analytes where analytical CV is greater than allowable CV, the established RCV may be broader and may fail to detect true positives.
- Large number of actual patient test results need to be checked over a longer duration of time.
- Manual verification for the detection of false negatives was not done.
Financial Support and Sponsorship
Nil.
Conflicts of Interest
This study was presented as a poster in the South Regional Conference of the Association of Clinical Biochemists of India 2014.
| References | | |
| 1. | |
| 2. | Pearlman ES, Bilello L, Stauffer J, Kamarinos A, Miele R, Wolforet M. Implications of autoverification for the clinical laboratory. Clin Leadersh Manag Rev 2002;16:237-9.  |
| 3. | Felder R. Laboratory reporting for the future: Linking autoverification to the electronic medical record. American Association for Clinical Chemistry. AACC Continuing Education. Available from: http://www.aacc.org/access/record/default.stm. [Last accessed on 2014 Feb]. |
| 4. | Shih MC, Chang MH, Tien N, Hsiao CT, Peng CT. Building and validating an autoverification system in the clinical chemistry laboratory. Lab Med 2011;42:668-73. |
| 5. | Burtis CA, Ashwood ER, Bruns DE. Tietz Textbook of Clinical Chemistry and Molecular Diagnostics. 4 th ed. Missouri: Elsevier Saunders; 2006. p. 467-71. |
| 6. | Clinical and Laboratory Standards Institute. Autoverification of Clinical Laboratory Test Results; Approved Guideline. (CLSI document AUTO10-A). Wayne, Pennsylvania: Clinical and Laboratory Standards Institute; 2006. |
| 7. | Kim WJ, Kim QJ, Kim SI. Differential application of rate and delta check on selected clinical chemistry tests. J Korean Med Sci 1990;5:189-95. |
| 8. | Ricos C, Alvarez V, Cava F, Garcia-Lario JV, Hernandez A, Jimenez CV, et al. Desirable specification for total error, imprecision, bias, derived from intra and inter-individual biological variation. Scand J Clin Lab Invest 1999;59:491-500. |
| 9. | Park HS, Kim SY, Lee W, Chun S, Min KW. New decision criteria for selecting delta check Methods based on the ratio of the delta difference to the width of the reference range can be generally applicable for each clinical chemistry test item. Ann Lab Med 2012;32:345-54. |
[Table 1], [Table 2], [Table 3]
| This article has been cited by | | 1 |
Designing and validating an autoverification system of biochemical test results in Hatay Mustafa Kemal University, clinical laboratory |
|
| Bahar Ünlü Gül, Oguzhan Özcan, Serdar Dogan, Abdullah Arpaci | | Biochemia Medica. 2022; 32(3) | | [Pubmed] | [DOI] | | | 2 |
Application and optimization of reference change values for Delta Checks in clinical laboratory |
|
| Jinyoung Hong,Eun-Jung Cho,Hyun-Ki Kim,Woochang Lee,Sail Chun,Won-Ki Min | | Journal of Clinical Laboratory Analysis. 2020; | | [Pubmed] | [DOI] | | | 3 |
Autoverification of test results in the core clinical laboratory |
|
| Edward W. Randell,Sedef Yenice,Aye Aye Khine Wamono,Matthias Orth | | Clinical Biochemistry. 2019; | | [Pubmed] | [DOI] | |
|
 |
|
|
|
Search Pubmed for