1. It should be verified that hospitals providing higher results express these as pg/mL, rather than pmol/L. It is not
unusual that in multicenter studies some participant forgets conversion from his/her routine working units into the agreed units. Indeed, the striking differences between the three hospitals disappear if the series of highest results
is divided by 3.43, the conversion factor to pg/mL.
2. The goodness of results reported by all three hospitals may be shown by the very close correspondence of ratio values between the methods: for all three hospitals, the Immulite/Elecsys ratio is 1.02 ± 0.11, and the Abbott/Immulite ratio 1.29 ± 0.01. Results are therefore reliable, at least «relatively».
3. The wide variability between methods (advocated as invalidating of K/DOQI references) is not such in practice, because 90% of hospitals use the Immulite (from DIPESA) and the Elecsys or Modular (from ROCHE) systems, and as we have just seen, the results of both are «exchangeable ».
4. It is a fact known to all laboratory technicians that differences in PTH do not only stem from the existence of various circulating parathyroid peptides and different methods to measure them, but also from a characteristic
physical property of such peptides, namely their extreme lability to time and temperature.2 The only «real» values are those reported during surgical procedures, when rapid measurements are made under no temperature changes. The situation is quite different when serum is not separated until 1 hour after blood collection, is kept in test tubes in racks for 2 additional hours at room temperature, subsequently frozen for one week, and finally thawed and let to reach room temperature for 1 or 2 additional hours: PTH levels are up to 30% lower than those found with a rapid measurement.3 This may be the reason for the different results in the hospital with the lowest values as compared to the other two hospitals.
It would thus be desirable that nephrologists become interested not only on the methods used to measure PTH, but also on some pre-test and sample traceability issues having an impact on the quality of results, such as whether or not the processed sample was frozen, and whether the laboratory working system allows for delays (times longer than 40 min) in the serum separation (serum is the sample of choice for measuring PTH), freezing, and thawing phases. We technicians are fully aware of the need for rapid serum separation, freezing, and testing, but not all laboratories have automated sample management systems for prioritizing, for instance, PTH samples and minimizing duration of processing steps. However, this will soon occur, and will result in the desired decrease in
differences between PTH measurements.