Ethnic disparities in colon cancer survival in New Zealand Dr Diana Sarfati University of Otago Wellington Māori Cancer Conference; Aug 2009
Acknowledgements Sarah Hill Project team Bridget Robson, Donna Cormack, Tony Blakely, Gordon Purdie, Liz Dennett, Kevin Dew Cancer Society of New Zealand
Colorectal cancer mortality rates by ethnicity Blakely T, Tobias M et al. Tracking disparity: Trends in ethnic and socioeconomic inequalities in mortality, 1981-2004. Wellington: Ministry of Health
Colorectal cancer incidence by ethnicity for men aged 25 yrs + Shaw, Blakely, Atkinson, Tobias, Sarfati, Cunningham. CancerTrends: Preliminary results. [Not for distribution]
Maori : non-maori/ non-pacific RSRR 1 Maori: non-maori/non-p Pacific RSRRs 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Bladder Breast Cervix Colon/rectum Lung Prostate Jeffreys M, Stevanovic V, et al. Ethnic differences in cancer survival in New Zealand: linkage study. Am J Public Health 2005;95:834-7.
Proportion attributable to stage Maori:non-Maori/non-Pa acific R SRRs 1.00 0.90 0.80 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00 0.05 0.03 0.1 0.07 0.23 0.08 0.38 0.23 0.47 0.07 0.03 0.92 0.1 0.82 0.7 0.7 0.59 0.43 Bladder Breast Cervix Colon/rectum Lung Prostate Age-standardised Additional stage standardisation Remaining portion Jeffreys M, Stevanovic V, et al. Ethnic differences in cancer survival in New Zealand: linkage study. Am J Public Health 2005;95:834-7.
Study Questions: Verify Māori / non-māori survival disparity (for colon ca) What factors contribute to this survival disparity? Tumour characteristics Patient comorbidity Health services
Retrospective cohort: 301 Māori, 328 non-māori patients diagnosed 1996-2003 NZ Health Information Service Cancer Registry Study Cohort Ethnicity Hospitalisation Database Medical Charts Demographics Tumour Comorbidity Health care Mortality Database Survival
Cancer-specific survival Surviva al
Reasons for poorer survival in Māori???? More aggressive / advanced tumours Greater comorbidity Poorer health care Differences in treatment Differences in markers of health care access / quality
Tumour characteristics: grade of tumour 80 70 60 Well differentiated Moderately differentiated Poorly differentiated Percentageo of cohort 50 40 30 20 10 0 Mäori non-mäori
Tumour characteristics: stage at diagnosis* Percentage of cohort 40 35 30 25 20 15 Stage I Stage II Stage III Stage IV Unstaged 10 5 0 Māori non-māori *Age- and sex-standardised prevalence
χ Reasons for poorer survival in Māori?? More aggressive / advanced tumours Greater comorbidity? Poorer health care Differences in treatment Differences in markers of health care access / quality
Patient comorbidity* 60 p=0.0001 Māori non-māori 50 Percentage e of cohort 40 30 20 P<0.0001 P<0.0001 P=0.02 P=0.008 10 0 Hypertension Previous heart attack Heart failure Diabetes Respiratory disease Previous stroke or TIA Renal disease Neurological disease *Age- and sex-standardised prevalence
Effect of comorbidity on survival Adjusted hazard ratios for cancer specific and all-cause survival among 589 colon cancer patients by specified comorbidity 4.5 4 3.5 3 2.5 2 1.5 Cancer specific all-cause Hazard Ratio 1 0.5 0 Angina Hypertension Previous MI Arrhythmias CHF PVD Resp. disease GI ulcer Other cancer Cerebrovasc disease Diabetes Renal disease Other neuro 0 1 0r 2 3 or more 0 1 2 3 or more Charlson Index Comorbidity count Adjusted for age, sex, ethnicity, smoking, year of diagnosis, stage, grade, site of cancer
Effect on treatment choice Of 190 patients in our cohort with Stage III disease, 68% were offered chemotherapy. Older patients and those with higher comorbidity were considerably less likely to be offered chemotherapy. 84% with Charlson comorbidity score=0 cf 19% with Charlson comorbidity score of 3+ were offered chemotherapy 80% of 55-64 yr cf 37% of 75yr + were offered chemotherapy Among those with highest comorbidity there was around a 60% reduction in excess risk of death if offered chemotherapy.
Reasons for poorer survival in Māori?? χ More aggressive / advanced tumours Greater comorbidity Poorer health care Differences in treatment Differences in markers of health care access / quality
100 90 80 Treatment Surgery* Māori non-māori Percentage e of cohort 70 60 50 40 30 20 10 0 Removal primary tumour Palliative operation Elective surgery *Age- and sex-standardised prevalence Emergency surgery Obstructed Perforated Obstructed or perforated
Treatment: number of lymph nodes removed during surgery 80 70 60 percentage of co ohort 50 40 30 Māori non-māori 20 10 0 0-11 12-29 30+ Number of lymph nodes
Post-operative mortality 20 Māori non-māori Adjusted for patient and clinical factors: Percentage of cohort 15 10 5 All surgery: RR = 3.17 (1.51-6.63) Elective surgery: RR = 5.15 (1.37-19.28) 0 Death following any surgery Death following elective surgery
100 90 80 Chemotherapy (stage III)* Treatment Referred to oncologist Reviewed by oncologist Offered adjuvant chemo Percentage of coh hort (stage III) 70 60 50 40 30 20 Received adjuvant chemo Started within 8 weeks 10 0 Māori *Age- and sex-standardised prevalence non-māori
Markers of health care access / quality Proportion of (Māori/n non-māori) cohort 70 60 50 40 30 20 10 Treatment facility type 50 40 30 20 10 Area deprivation Māori non-māori 100 80 60 40 20 Rurality 0 2 public 3 public private 0 1 (high SES) 2 3 4 5 (low SES) 0 Urban Semiurban Rural
Reasons for poorer survival in Māori? χ More aggressive / advanced tumours Greater comorbidity Poorer health care Differences in treatment Differences in markers of health care access / quality
Māori/non-Māori hazard ratio (RR): step-wise adjustment for explanatory variables Adjusted for: HR (95% CI) Unadjusted 1.33 (1.03-1.71) i) Demographics 1.30 (0.99-1.71) ii) + Tumour characteristics 1.33 (0.99-1.79) iii) + Patient comorbidity/smoking 1.20* (0.89-1.63) iv) + Treatment 1.17 (0.86-1.60) v) + Health care access / quality 1.07* (0.77-1.47) *Signficant change from previous HR (Hausman test) Hausman J. Specification tests in econometrics. Econometrica 1978;46(6):1251-7
Conclusions Māori patients have significantly poorer survival from colon cancer compared with non-māori patients (HR 1.33) Greater comorbidity levels and differences in health care access are both important mediators of worse survival in Māori (each accounts for ~1/3 of the total disparity)
What does this mean? Māori patients have poorer access to quality health care compared with non-māori Māori patients have lower cancer survival compared with non-māori
Where to from here? System factors Resourcing and location of cancer services Focus of cancer services (structure, organisation delivery of services reflect Pakeha world view) Composition of cancer service workforce Regional factors Improve access to specialists in rural areas Increase support for patients and whãnau travelling to cancer services Specialist support for local clinicians Coordination of case management through cancer care pathway Clinical factors Optimise treatment of those with comorbidity Evaluation of patient management against clinical guidelines/ audit in peer review context Training in cultural safety