Peripheral Blood Smear Utilization within a Military Hematology/Oncology Fellowship Program

Background: The practice of peripheral blood smear (PBS) interpretation is considered an important diagnostic skill in Hematology/Oncology (HO). The Accreditation Council of Graduate Medical Education (ACGME) requires trainees to become proficient in the preparation and interpretation of blood smears. We explored practice patterns of PBS utilization in a tertiary care, military medical center HO fellowship program.

Methods: A retrospective chart review was conducted on 350 consecutive outpatient referrals for benign hematology diagnoses (ICD-9 codes 280-289) in a military medical center, from August 2011 to March 2012. Data obtained from the electronic medical records (EMR) included patient demographics, pertinent medical history, labs ordered, and provider status (resident, fellow, or staff). We also recorded HO ordering patterns for complete blood counts (CBC) and PBS and documentation of PBS findings. For those encounters in which the findings were documented, we reported any further diagnostic studies ordered. The practices of PBS review by HO clinicians and hematopathologists (HP) were also compared.

Results: 334 of 350 records reviewed met inclusion criteria for the study. Of the 16 records excluded, one was due to age > 85, 11 due to malignant oncology diagnoses, and 4 were unavailable within the EMR. The median age was 52.5 with males representing 52% of the cohort. 43.8% were self-reported as Caucasian, 23.6% African-American, and 32.6% “Other.” The distribution of diagnoses were 52.7% (n = 176) erythrocyte (RBC) disorders, 23.0% (n = 77) leukocyte (WBC) disorders, 10.2% (n = 34) platelet (PLT) disorders, and 14.1% (n = 47) falling into the category of “other.” The five most common diagnoses were anemia, unspecified (n = 74, 19.3%), iron deficiency anemia (n = 53, 13.8%), leukocytosis, unspecified (n = 37, 9.6%), thrombocytopenia, unspecified (n = 31, 8.1%), and leukopenia, unspecified (n = 16, 4.2%) corresponding to ICD-9 codes 285.9, 280.9, 288.6, 287.5, and 288.5, respectively.

A CBC was ordered in 82.9% encounters (n = 277/334, 95% confidence interval [CI], 78.9%-87%) A PBS was ordered in 52.3% of cases in which a CBC was ordered (n = 145/277, 95% CI, p < 0.001). There was no statistical significance between the rates at which residents, fellows, or staff ordered PBS with percentages of 40% (n = 18/45), 43.6% (n = 89/204), and 44.7% (n = 38/85), respectively. A HP reviewed 5.7% (n=19/334) of the CBC’s obtained.

If a PBS was ordered, findings were documented in the EMR for 49% of cases (71/145). In the 71 encounters that had PBS documentation, only 7 had HP review (9.9%). Examination of PBS led to further evaluation in 40.8% (29/71) of cases, which represented 8.7% (29/334) of the entire cohort. Of these 29, there was one consult to Gastroenterology, one imaging study ordered (ultrasound for splenomegaly), 9 bone marrow biopsies (12.7%), and lab work in all 29. Of the labs ordered, 8 (27.6%) were studies for hemolysis (DAT, LDH, haptoglobin), 10 (34.5%) were studies for myeloproliferative disease (Jak2, BCR-ABL), and 5 (17.2%) were peripheral flow cytometry.

A PBS was more likely to be ordered for WBC disorders (71.9%, CI 95%: 61.9%-81.9%; p ≤ 0.001), followed by PLT (61.3%, CI 95%: 44.2%-78.4%; p = 0.076) and RBC disorders (38.7%, CI 95%: 30.9%-46.5%; p ≤ 0.001), respectively. Of the 29 cases in which PBS findings were documented and further studies ordered, 8 (27.6%) were RBC disorders, 10 (34.5%) were WBC disorders, 6 (20.7%) were PLT disorders, and 5 (17.2%) fell into the category of “other.” In the “other” category, 2 were other specified disease of blood and blood-forming organs (289.89), and the remaining 3 were splenomegaly (789.2) polyclonal hypergammaglobulinemia (273), and monoclonal paraproteinemia (273.1).

Discussion: The practice of interpreting the PBS is considered an important diagnostic skill for HO providers. In a military HO training program, PBS were ordered in about half of new benign hematologic outpatient referrals. When PBS were ordered, findings were documented in 49% of cases. While diagnoses could be made without getting PBS in a majority of cases, close to 10% of all benign hematology referrals had documented PBS fidings that led to further diagnostic evaluation.