25803630

PMC4828630

CC999999/Intramural CDC HHS/United States

CDC Public Access

Introduction

The results of a 2001–2005 polycythemia vera (PV) investigation in Eastern Pennsylvania revealed a disease cluster plus underreporting and false reporting to the Pennsylvania Cancer Registry (PCR).

Purpose

The objectives of this study were 1) to assess PV reporting to the PCR in 2006–2009, 2) to determine whether a cancer cluster persisted, and 3) to determine whether other myeloproliferative neoplasms (MPNs), including essential thrombocytopenia (ET), were subject to similar reporting problems.

Methods

Cases were identified from: 1) PCR records from the Tri-County, 2) reviewing billing records at Tri-County hematologist/oncologist offices, and 3) self-identification. An expert panel of physicians reviewed medical records and determined “true,” “false,” or “indeterminate” cases reported to the PCR. The analyses were conducted to determine sensitivity and positive predictive value (PPV) of case reporting to the PCR, estimate cancer incidence rates, and evaluate the presence of cancer clusters.

Results

Of 290 cases identified, 90% were from the original PCR, 9% from billing records, and 1% from self-report. Fifty-five cases consented to participate, and medical records were obtained for 44. The expert panel determined that 45% were true cases, 32% were false cases, and 23% were indeterminate. PV had 100% (95% CI, 59–100) sensitivity, but only 47% PPV (95% CI, 20–70): ET had 78% (95% CI, 47–99) sensitivity and 100% PPV (95% CI, 59–100). Low participation and chart review rates led to rates with wide confidence intervals. We did not identify any PV cancer clusters, but we did identify a cluster of 9 ET cases in the Wilkes-Barre, Pennsylvania area.

Conclusion

The current study was limited by the low response rate (22%) from MPN patients in the Tri-County area. This study identified 47% PPV for PV reporting and 100% PPV for ET.