Quality assurance for multiplexed assays - how can it be achieved?

Multiplexed assays are now a common form of analysis in routine clinical and research laboratories. Assuring the quality of this type of complex, massively-parallel testing poses challenges not encountered in the traditional single-plex assay. A range of quality assurance measures is implemented at different stages in a multiplex assay, beginning in the manufacturing process, and the ensuing analytical and data interpretation stages. This article explores quality issues and the quality assurance measures that have been devised for multiplex assays ranging from simple two-plex assays to the types of assays that involve simultaneous testing on millions of test sites on a single analytical device.

Early pregnancy diagnosis in swine by direct radioimmunoassay for progesterone in blood spotted on filter paper.

A direct radioimmunoassay method for the measurement of progesterone in blood dried on filter paper has been developed for the early pregnancy diagnosis in sows, as well as for monitoring progesterone levels during the oestrous cycle. Pregnancy diagnosis was performed with 95 sows on Days 17-22 after artificial insemination (AI). The cut-off value for pregnancy diagnosis of 7.5 ng/ml was calculated (mean+/-2S.D.) from the progesterone concentrations measured on the same days from non-inseminated animals. There were 85 cases considered pregnant on the basis of progesterone concentration, leaving 10 animals non-pregnant. The accuracy for the positive cases was 98.8%. Two of the 10 sows considered as negative subsequently farrowed, giving an accuracy of 80%. The overall accuracy of the method was 96.8%. The blood-spot assay may be a useful tool for early pregnancy diagnosis in swine, with respect to sampling, simplicity, speed and accuracy.

[Evaluation of home pregnancy tests]. / Vurdering af graviditetstest til hjemmebrug.

Three home pregnancy tests, Clearblue (Tjellesen), Clearblue one step (Tjellesen) and Predictor (Organon), were tested and compared with respect to sensitivity, specificity, prozone-phenomenon and convenience for use. In addition, they were compared with two well-known pregnancy tests, HCG-Novoclone (Novo) and HCG-Nostick (Organon), which are used in several hospitals and drugstores. All three home pregnancy tests were found to be in agreement with the instructions for use with regard to sensitivity and prozone phenomenon. One test, Clearblue one step, showed false positive results when LH was added. Clearblue was found to be laborious and unhygienic in performance. Predictor seems to be the test of choice.

[Clinical trial of 2 highly sensitive pregnancy tests--Tandem ICON HCG-urine and OPCO One-step Pacific Biotech]. / Klinisk afprøvning af to højsensitive graviditetsprøver--Tandem ICON HCG-urine og OPCO "One-step" Pacific Biotech.

The highly sensitive pregnancy test, Tandem ICON HCG-urine, was compared with OPCO "One-step" Pacific Biotech in the gynecological and obstetrical department of Randers Hospital. The purpose was to determine the test specificity and sensitivity of the pregnancy tests in women with possible pregnancies. The test sensitivities were found to be 0.83 and 0.63 for ICON and ONE, respectively. We experienced one and four cases of negative pregnancy tests and ectopic pregnancies with ICON and ONE, respectively. In all cases, the diagnosis of ectopic pregnancy was suspected by other means and the patients submitted to operation. The ICON test showed its supremacy in diagnosis while ONE was easier and cheaper.

Early detection of ectopic pregnancy. Use of a sensitive urine pregnancy test and transvaginal ultrasonography.

We performed a prospective study to evaluate the reliability of a rapid monoclonal antibody urine pregnancy test with a sensitivity limit of 20 mIU/mL combined with transvaginal ultrasonography in the early diagnosis of ectopic pregnancy in 116 women with subacute pelvic pain and a stable general condition. The diagnosis of tubal pregnancy was confirmed with laparoscopy in 100 of the 103 women with positive sensitive urine pregnancy tests and no intrauterine gestational sac at transvaginal ultrasonography. Laparoscopy revealed a hemorrhagic corpus luteum in four of the eight subjects with negative monoclonal antibody pregnancy tests and no intrauterine gestational sac, an ovarian cyst in three and a normal pelvis in one. Of the five women with a positive pregnancy test and an intrauterine gestational sac, two had a hemorrhagic corpus luteum, two a normal pelvis and one a tubal pregnancy at laparoscopy. The sensitivity of a monoclonal antibody urine pregnancy test and transvaginal ultrasonography combined for the diagnosis of ectopic pregnancy was 99%, and the specificity was 80%, with positive and negative predictive values of 97% and 92%, respectively.

The impact of an instant pregnancy test kit on the operations of a major hospital casualty department.

The records of all patients on whom a casualty department pregnancy test was performed during the first 4 months of 1986 were retrospectively examined and compared with the records of all patients who in the first 4 months of 1987 had had an instant pregnancy test performed in the same casualty department, to determine the impact of such a kit on diagnostic accuracy, operative procedures and related economic factors. It was demonstrated that instant pregnancy testing significantly improved the accuracy of provisional diagnoses and appeared to help in reducing the number of surgical procedures performed. Furthermore, it was shown that a combination of careful clinical assessment plus the appropriate application of an instant pregnancy test kit could result in a cost saving to the hospital of over $41,000 per annum.

PIP: To assess the impact of an instant pregnancy test kit on hospital admissions for suspected ectopic pregnancy, the records of 100 patients from a Sydney, Australia hospital were examined. All records of patients of patients whom a pregnancy test was performed during 4 months of 1986 were compared with records of patients who had had an instant pregnancy test performed in the same hospital during 4 months of 1987. 2 test methods were used, the Tandem B HCG RIA which takes 2 hours, and the instant Test Pack HCG serum test kit, which takes 3 minutes. There was a 100% accuracy correlation between the 2 tests. Costs were compared on the basis of how many patients were admitted to the hospital who might not have required admission had a test result been available. Instant pregnancy testing significantly improved the accuracy of provisional diagnosis and seemed to contribute to a reduction in the number of surgical procedures performed. Calculations using the number of hospital admissions that would not have been made had a test result been available, revealed estimated savings of $41,000/per year to the hospital. It was concluded that the Test Pack was cheap, easy to use, provided instantaneous results, and resulted in significant savings in hospital operating costs.

Performance and sensitivity of modern home pregnancy tests.

We have studied the performance of three currently available home pregnancy tests. The Advance is a test based on monoclonal antibodies in an enzyme immunoassay format that specifically detects hCG in urine. The positive results are determined by the presence of a blue color in a color stick (30 minutes). The Daisy 2 and FACT are tests based on monoclonal antibodies that specifically detect hCG in urine in a hemagglutination inhibition assay; positive results are demonstrated by the deposition of a dot ring at the bottom of the test tube (45 minutes). Thirty-five patients who eventually became pregnant in that cycle collected the first morning urine specimens from day of ovulation (determined by ultrasound) in artificial insemination donor (AID) cycles, or from the date of gamete intrafallopian transfer (GIFT) for 16 consecutive days. In each specimen, the Advance, Daisy 2, and FACT tests were performed; in addition, beta-hCG levels were determined by radioimmunoassay. All the home pregnancy tests studied can detect pregnancy as early as 9 or 10 days post-conception, and they give positive results at the time of the expected onset of menses in 70%, 95%, and 88% of cases for Daisy 2, Advance, and FACT, respectively. The sensitivity of the these home pregnancy tests was determined to be about 200 mIU/mL of urine after correlating their positive or negative results with the concentrations of urinary beta-hCG determined by radioimmunoassay.

The effect of different human chorionic gonadotropin assay sensitivity on screening for ectopic pregnancy.

The distribution of human chorionic gonadotropin levels in 184 patients with ectopic pregnancy is examined. The impact of changing the sensitivity of human chorionic gonadotropin testing on the incidence of false negative results is discussed. Pregnancy blood tests with a detection threshold of 200 mIU/ml were associated with an 11.9% incidence of false negative results.

Evaluation of the beta-Neocept test for early pregnancy.

We tested the validity of the beta-Neocept test for early pregnancy against that of the plasma human chorionic gonadotropin beta-subunit radioimmunoassay (beta-HCG RIA). The beta-Neocept test had a sensitivity of 88%, a specificity of 93%, a positive predictive value of 95%, a negative predictive value of 84% and an accuracy of 90%. In view of these performance characteristics, its low cost and its ease of use, the beta-Neocept test could be used as the initial pregnancy test when there is a high probability of pregnancy, as there was in this study population, which consisted of 111 women attending endocrine infertility clinics. The more expensive beta-HCG RIA could be reserved for special indications and for patients in whom the results of the urinary hemagglutination inhibition tests are inconsistent with the clinical signs and symptoms.

PIP: Paired urine and blood samples from 111 women attending endocrine infertility clinics in Vancouver, Canada, were used to compare the validity of the beta-Neocept test for early pregnancy against that of the plasma human chorionic gonadotropin beta-subunit radioimmunoassay (beta-HCG-RIA). None of the urine specimens contained glucose, 1 specimen contained 1 + protein measured by dipstick, and 1 contained a large amount of blood. 67 patients were pregnant and 44 were not pregnant. Beta-Neocept test results were positive for 59 pregnant patients, negative for 8 pregnant patients, positive for 3 nonpregnant patients, and negative for 41 nonpregnant patients. There were 62 positive and 49 negative tests. Results with the Beta-Neocept test became positive as early as the 28th day after the onset of the last menstrual period and negative results occurred with plasma beta-HCG levels ranging between 5 and 157 IU/1. The sensitivity of the Beta-Neocept test was 88%, the specificity 93%, the positive predictive value 95%, the negative predictive value 84%, and the overall accuracy 90%. The 8 false-negative results all occurred at low plasma beta-HCG levels. The false-negatives occurred in 1 patient on no medication, 1 taking clomiphene citrate and prednisone, 1 taking gonadotropin releasing hormone (GRH), 1 taking human menopausal gonadotropin and human chorionic gonadotropin, 1 taking GRH and bromocriptine mesylate, 1 taking clomiphene citrate alone, and 2 taking clomiphene citrate and progesterone vaginal inserts. The 3 false-positive results occurred in 2 patients using progesterone vaginal inserts for luteal-phase support and 1 who was taking clomiphene citrate. The results indicate that the Beta-Neocept test is sensitive and specific, and would be an appropriate initial test for patients with a high probability of pregnancy. The beta-HCG-RIA could be reserved for patients in whom the results of the less expensive and less difficult Beta-Neocept were inconsistent with clinical signs and symptoms.

Early urinary pregnancy testing. Correlation with serum beta-HCG radioimmunoassay.

Urinary pregnancy testing with Sensitex, a tube macroflocculation system, produced a positive test by 17 days after basal-body-temperature-timed ovulation in 59% of determinations. Data from days 18 to 21 showed that a negative test was associated with a statistically significantly increased propensity for first-trimester abortion. Freezing of the urinary specimens, but not refrigerated sm, produced a positive test by 17 days after basal-body-temperature-timed ovulation in 59% of determinations. Data from days 18 to 21 showed that a negative test was associated with a statistically significantly increased propensity for first-trimester abortion. Freezing of the urinary specimens, but not refrigerated storage for five days, adversely affected the accuracy of the test.

Heterophilic antibodies interfering with radioimmunoassay. A false-positive pregnancy test.

A young man with amenorrhea had a consistently positive pregnancy test result (serum radioimmunoassay measurement of beta-human chorionic gonadotropin hormone). No fetal or placental tissue was found after uterine curettage and exploratory laparotomy. The false-positive pregnancy test result was due to heterophilic antibovine and antigoat antibodies in the patient's serum. These antibodies interfered with radioimmunoassays using goat antibodies. This case shows that serum heterophilic antibodies can interfere with immunoassays and result in unnecessary diagnostic procedures and/or unnecessary treatment.

Procedures for measuring accuracy and sensitivity of immunochemical pregnancy test kits.

Analytical procedures were used to measure the accuracy and sensitivity of immunochemical pregnancy test kits. Performances of all currently marketed hemagglutination inhibition, latex agglutination inhibition, and direct latex agglutination pregnancy kits were evaluated.

PIP: 7 immunochemical methods are currently available for measuring human chorionic gonadotropin (HCG) which is secreted during pregnancy, reaching its highest level during the 1st trimester: 1) latex agglutination inhibition (LAI), 2) hemagglutination inhibition (HAI), 3) direct latex agglutination (DLA), and 4) radioimmunoassay (RIA). In a study of 5 commercial immunochemical pregnancy kits, it was concluded that the HAI kits were more accurate and sensitive than the LAI or DLA kits. The need for a standardized, transferrable analytical method for evaluating kit performance was expressed by the Food and Drug Administration and by increased over-the-counter sales of these products. This paper presents protocols developed for testing the performance of these kits including the following aspects: accuracy and sensitivity determination, test procedures for HAI, LAI, and DLA, apparatus, and reagents. A test was performed on 206 assorted lots of kits with the following results: 1) none of the 98 HAI kits gave false positive or false negative results, 2) the LAI kits gave 24 false and 3 inconclusive results, 3) the mean level of HCG detected by HAI kits was below that obtained with LAI and DLA kits; thus the HAI detects pregnancy at an earlier stage, 4) average sensitivity of all kits tested was greater than or equal to that claimed by the manufacturer, 5) denaturation of the antisera may enhance the sensitivity and given an increase of false positive reactions, 6) low concentrations or HCG standard in dilute solutions were unstable due to absorption of HCG on glass surfaces, 7) erroneous results can be caused by HCG secreted in hydatidiform moles, choriocarcinoma and certain tumors, ectopic pregnancy, and protein in the urine, 8) data from repeat analyses could be misleading, and 9) reagents should be visually inspected before use to detect bad reagents.