Quality Issue: Marrow Interpretations

To the Editor:

In the past, hematologists/oncologists performed bone marrows and often forwarded clot and biopsy material (in some instances, aspirate samples as well) to local laboratories. In this manner, the hematologist/oncologist was able to review the slides personally or with the pathologist. This process permitted the correlation of clinical and pathologic information and allowed for more accurate diagnoses upon which treatment decisions were made. The time for this entire process was brief (several days).

In the current cost conscious climate, insurance companies and managed care organizations arrange the processing of pathology samples and subsequent interpretations by reference laboratories. Oftentimes, these laboratories are no longer next door, nor are the clinicians able to relate personally with slides or pathologists. We decided to measure the accuracy and timeliness of bone marrow samples (performed in our outpatient clinic) that were forwarded to reference laboratories.

Bone marrow clot and biopsy samples obtained on 28 patients over 6 months at the Baptist Regional Cancer Center were logged for rapidity of returned reports, frequency of returned slides, and accuracy of pathology interpretation (Table 1). In each case, a request was made for slides to be returned for personal review. Four reference laboratories were used, as noted by anonymous letter code. Bone marrow verbal or written reports were received at a median of 9 days (range, 2 to 28 days). Only 11 of 28 (39%) requests resulted in returned slides; median time for receipt of slides was 21 days. I personally reviewed all returned slides; however, when disagreement occurred, I asked one of our next-door pathologists to re-review the slides in an impartial manner without indicating I was assessing a potential quality issue. Errors were noted in 5 of the 11 (45%) cases. No attempt was made to assess the error frequency rate for each reference laboratory because of the limited sample size; but, of the five incorrect diagnoses, they occurred from three of the four labs that were used, as shown in Table 1, item 2.

The purpose of this letter is to sound an alarm! In the past, we had quick turnaround and an opportunity to review slides in the presence of our pathologists. Patients and family received reports quickly and we were able to react promptly with treatment decisions. Now, there is a considerable delay and many times we do not have the chance to review slides. Errors are being made when pathologists lack clinical input. Some of my fellow hematologists/oncologists have become complacent because of the difficulty in reviewing slides; unfortunately, they are no longer asking for them. The current environment has placed the diagnoses in the eyes of pathologists who are remote from the clinical material and about whom we know little of their expertise.