Building a Population-Based Cancer Registry in Dar es Salaam, Tanzania: A Pilot Project to Evaluate Completeness and Validity of Pathologic Data
Cancer remains an under-recognized health condition throughout most of Africa, and improved surveillance systems for determining cancer incidence, mortality, and prevalence of risk factors are badly needed. HIV and infectious diseases remain prevalent in Africa, with as many as 36 percent of cancers in Africa being infection-related — twice the world average. Increases in HIV-induced immunosuppression have resulted in an increased risk of many cancers, while economic development and changing exposures to common carcinogens have been associated with increases in the prevalence of cancers of the lung, breast, esophagus, and prostate. Currently, there is widespread uncertainty about the role of various putative risk factors for the development of cancers in Africa, and thus, the most advantageous targets for cancer prevention initiatives are unknown. This pilot study forms part of a larger initiative to rebuild a population-based cancer registry in Tanzania.
- Aim #1: To evaluate the completeness of cancer registry data collection utilizing innovative technology for remote data entry of pathologically diagnosed malignancies in Dar es Salaam, Tanzania.
- Aim #2: To evaluate the validity of pathologic diagnoses of HIV-associated malignancies in Dar es Salaam, Tanzania.
A 12-month pilot study will be divided into four quarters: (1) planning; (2) data collection; (3) pathology review and registry audit; and (4) data analysis, interpretation, and reporting of results. During the planning phase of the study, we will collaborate with the Global Health Research Foundation to implement novel technology for remote data entry using cellular modems to transmit data to a central registry database. The eight pathology departments where tumor specimens are reviewed throughout Dar es Salaam will be trained to utilize the remote data entry technology to report new pathologic diagnoses of malignancy. In the data collection phase, we will aim to document all pathologic diagnoses of malignancy from these eight pathology departments in a central registry. The registry will utilize CanReg5©, which is an opensource software package produced by the International Agency for Research on Cancer.
Following the pilot period of data collection, we propose to measure two principles which have previously been established as critical to cancer registry data quality: (1) Completeness of reporting to the cancer registry; and (2) Validity of pathologic interpretations of malignancies. In our assessment of validity, we will specifically focus on the HIV-associated cancers (Kaposi's sarcoma, lymphoma, and cervical cancer). In the analysis phase, we will conduct an audit of pathology records at the contributing sites to evaluate the completeness of the data submitted to the central registry database. To measure validity of the pathologic diagnoses, tissue slides from all diagnoses of HIV-associated malignancies as well as specimens obtained from patients known to be HIV-seropositive will be transported to the United States for review by Dr. Ann Nelson, an international expert in AIDS pathology. The accuracy of malignant diagnoses made by Dar es Salaam pathologists will be measured against the gold standard diagnosis made by Dr. Nelson. Demonstration of completeness of reporting pathologic diagnoses to a central registry database utilizing novel remote data entry technology will represent an advance in the management of cancer registries in resource-poor countries. Demonstration of the validity of pathologic interpretations of specimens is a critical component in efforts to gain international support and funding for the development and expansion of the registry.