NCCN guideline-concordant cancer care in sub-Saharan Africa: a population-based multicountry study of 5 cancers

Abstract

Background: To assess population-based quality of cancer care in sub-Saharan Africa and to identify specific gaps and joint opportunities, we assessed concordance of diagnostics and treatments with National Comprehensive Cancer Network Harmonized Guidelines for leading cancer types in 10 countries.

Methods: Adult patients with female breast cancer, cervical cancer, colorectal cancer, non-Hodgkin lymphoma, and prostate cancer were randomly drawn from 11 population-based cancer registries. Guideline concordance of diagnostics and treatment was assessed using clinical records. In a subcohort of 906 patients with potentially curable cancer (stage I-III breast cancer, cervical cancer, colorectal cancer, prostate cancer, aggressive non-Hodgkin lymphoma [any stage]) and documentation for more than 1 month after diagnosis, we estimated factors associated with guideline-concordant treatment or minor deviations.

Results: Diagnostic information based on guidelines was complete for 1030 (31.7%) of a total of 3246 patients included. In the subcohort with curable cancer, guideline-concordant treatment was documented in 374 (41.3%, corresponding to 11.7% of 3246 patients included in the population-based cohort): aggressive non-Hodgkin lymphoma (59.8%/9.1% population based), breast cancer (54.5%/19.0%), prostate cancer (39.0%/6.1%), colorectal cancer (33.9%/9.5%), and cervical cancer (27.8%/11.6%). Guideline-concordant treatment was most frequent in Namibia (73.1% of the curable cancer subcohort/32.8% population based) and lowest in Kampala, Uganda (13.5%/3.1%). Guideline-concordant treatment was negatively associated with poor ECOG-ACRIN performance status, locally advanced disease stage, origin from low Human Development Index countries, and a diagnosis of colorectal cancer or cervical cancer.

Conclusions: The quality of diagnostic workup and treatment showed major deficits, with considerable disparities among countries and cancer types. Improved diagnostic services are necessary to increase the share of curable cancer in sub-Saharan Africa. Treatment components within National Comprehensive Cancer Network Guidelines for several cancers should be prioritized.

Figure 1.

Flowchart of patients included in this study. Patients not traced: No information beyond cancer registry data available. For more information, refer to the detailed flowchart in Supplementary Figure 2 (available online).

Figure 2.

Treatment categories of patients adjusted for curability of disease, follow-up, staging and tracing in population-based cohort (n = 3246), stratified by cancer type. Filled bars: any cancer treatment documented; striped bars: no cancer treatment documented. Curable disease refers to stage I-III breast, cervix, colorectal and prostate cancer and aggressive Non-Hodgin lymphoma histopathological subtype, any stage. Noncurable disease refers to breast, cervix, colorectal and prostate cancer stage IV and indolent/unclassified Non-Hodgkin lymphoma. For details, see Supplementary Table 7 (available online). 30d = 30 days; CDT = cancer treatment; f-up = follow-up; NHL = Non-Hodgkin lymphoma.

Figure 3.

Guideline concordance (17,21-25) of treatment in patients with potentially curable cancer (n = 906), stratified by cancer type (for details, see Supplementary Table 6, available online). Evaluation of guideline concordance refers to Table 1 and Supplementary Table 3 (available online).

Figure 4.

A) Guideline concordance (17,21-25) of treatment in patients with potentially curable cancer (n = 906), stratified by registry (for details, see Supplementary Table 6, available online). Evaluation of guideline concordance refers to Table 1 and Supplementary Table 3 (available online). B) Guideline concordance of therapy for patients with potentially curable cancer in the population-based cohort (n = 3246), stratified by registry. Percentages refer to proportion of patients evaluated for guideline-concordance among all patients. Evaluation of guideline concordance refers to Table 1 and Supplementary Table 3 (available online).

Figure 5.

Factors associated with guideline-concordant treatment in patients with potentially curable cancer (n = 906, multivariable modified Poisson regression). Guideline-concordant treatment refers to combined guideline concordance or minor deviation from guidelines (17,21-25). Evaluation of guideline concordance refers to Table 1 and Supplementary Table 3 (available online). CI = confidence interval; DLBCL = diffuse large B-cell lymphoma.