Cancer epidemiology fieldwork in a resource-limited setting: Experience from the western Kenya ESCCAPE esophageal cancer case-control pilot study

Abstract

Background: Case-control studies remain an important study design for aetiologic research on cancer, particularly when cohorts are not available. In addition to the potential biases inherent in this design, conducting fieldwork in settings with weak health care and information systems for cancer, such as in sub Saharan Africa, confer additional challenges which we present here with the aim to share experience to guide future studies.

Methods: We undertook a hospital-based case-control study of squamous cell esophageal cancer at the Moi Teaching and Referral Hospital in Eldoret, West Kenya. Cases were recruited at endoscopy and controls from hospital wards, age and gender frequency-matched to cases. Urine, toenails, blood and tumour biopsy were collected and a questionnaire administered.

Results: During this pilot phase, 143 cases and 155 controls were successfully recruited. Complete questionnaire data was obtained through e-data collection. Biospecimen collection was possible with support of an already existing equipped laboratory. We introduce changes made in the main study phase, including on expansion of the control groups to allow to consideration of selection bias.

Conclusions: Extra attention and funding to train and monitor data quality and biospecimen collection and collaboration of a large group held together by strong leadership are essential. We recommend studies based on regional treatment centres with their more defined catchment areas rather than in the capital cities as referral routes in multi-level health care systems are severely attrition prone.

Keywords: Case-control study; Esophageal cancer; Fieldwork; LMICs.

Figure 1

The spatial distribution of geocoded residential locations of cases and controls, ESCCAPE Kenya esophageal cancer case-control study.

Figure 2

Examples of H&E-stained esophageal biopsies from patients in the study: (a) Low magnification (2×) of an entire biopsy, representative of a biopsy with a reasonable size for diagnosis; (b) The same biopsy with higher magnification (20×), revealing a well-differential SCC. Keratin pearls, as indicative of differentiation, are easily recognizable; (c) Moderately differentiated SCC, showing only single-cell keratinization (10×); (d) Poorly-differentiated SCC with no keratinization (10×) slides from esophageal cancer patients showing the different degrees of differentiation (Source: Dr Behnoush Abedi-Ardekani, Pathologist, IARC)

Figure 3:

Organogram of the local and international multidisciplinary team involved in the esophageal cancer case-control study fieldwork, indicating team members and their roles.