sequentially after examination. If visit of laboratory supervisor is not possible, the regional or provincial TB coordinator can select the slides on his or her planned supervisory visit. (b) Or the slides can be selected from those transported from PL to RL by mail or other measure. All the smear slides should be transported along with a list on which slide markings and microscopy results are clearly indicated and conform to the laboratory registry (or logbook). The reference laboratory supervisor or TB coordinator makes a list of smear slides selected and give them to a reference laboratory (or QA center) microscopist without the 1st reader's results in order to secure a blind-fashion evaluation. If any discrepancy is found when compared the results of 1st and 2nd readers, supervisor or TB coordinator ask another microscopist at RL or QA center to read them without knowing 1st and 2nd reader's results. The 3rd reader's results should be final and then final analysis is made before feed-back to the corresponding peripheral laboratory with an appropriate remedial actions to improve the quality of smear microscopy if necessary. Although the 2nd and 3rd readers cannot be absolutely reliable gold standard, there is no practical alternative for it.
This type of PT could be practiced twice in a year or maximum quarterly if it has been confirmed on the previous PT that the corresponding PL worker performed smear microscopy within acceptable standard. However it can be operated quarterly or even monthly if PL worker has been newly recruited or showed unacceptable proficiency on the previous PT. The frequency of PT and whether the slides will be selected from the slide boxes PL sent or on the spot during supervisory visit should be decided according to manpower and financial situation.
2] Slides keeping or sending: Peripheral laboratory must keep all the positive slides and part of negative slides after removing immersion oil with a few drops of xylene until supervisor or TB coordinator visits to collect them for QC or asks to send them. However it is desirable to keep no more than 250 negative slides from the last day of examination. The slides must be clearly and systematically labeled and kept positive and negative slides separately but in order consistent with the laboratory registry book (or logbook). Identification marking on slides must be able to cross-check the consistency of records on TB register in the clinic. The result of smear reading or name of patient should not be marked on slides. If PL send the slides, they have to send a list that must be filled with the slide labels and microscopy results in a consistent order of laboratory registry book.
3] Selection of smear slides: The number of slides for rechecking should be minimum enough to measure an acceptable sensitivity and specificity. It takes more time to determine a negative slide than to identify a positive one. Under the routine services a positive smear is found in approximately 5-20% of smear slides examining for diagnosis and treatment monitoring. In general it is recommended to select negative and positive slides samples separately.
The requiring sample size is determined in a way that the finding of a single positive slide among slides reported as negative will suffice for a decision. It must be possible to decide whether the laboratory procedures need review because the proportion of slides misclassified as negative is likely to exceed a critical value of 5%. A sample of slides may be slightly different depending on how many slides does a PL keep. If a PL keeps more than 200 slides, a sample of 41 negative slides might be a reasonable for external proficiency testing because it can rule out the failure to identify a laboratory which needs review of its operation with 95% certainty.(Table 1)
A positive slide is different from negative slide because real false positives are few and difficult to prove it due to a possibility of missing a few AFB on smear by 2nd and 3rd readers. Sometime it will harm good relationship between PL and RL. Rechecking positive slide should be focused on the staining property of AFB and correct grading of the results with a minimum number of slides.
TB coordinator or supervisor should cross-check the labels of smear slides with the laboratory logbook beforehand in order to confirm whether they kept all the slides and in a consistent order. And the smear check Forms 1 (see below) is filled up entering the laboratory serial numbers and reading results of the selected slides in the order of their serial placement in a slide box and then serial PT numbers are given by TB coordinator or supervisor. This from will be retained to the TB coordinator or supervisor. The Form 2 is a duplicate of Form 1 without information of the first reader's reading results and laboratory serial numbers. The Form 2 is placed inside the slide box for delivery to the rechecking microscopist.
4] Slide-checking: The medical technologist or microscopist of the reference laboratory (second reader) should assess smear preparation and staining prior to rechecking under microscope. The smears prepared too thin, too thick, too small, too large or sloughed off should be evaluated as "poor smear preparation" although it is utterly subjective evaluation. If the stained smear slides are kept at
