NAACCR / IACR Combined Annual Conference, Vancouver, Canada 9 - PowerPoint Presentation

Slide1

NAACCR / IACR Combined Annual Conference, Vancouver, Canada 9

th-13th June 2019

Socio-economic position and prevalence of comorbidity in cancer patients in England:a population-based study of four cancers

Helen Fowler, Aurelien Belot, Libby Ellis, Camille Maringe, Miguel Angel Luque-Fernandez, Edmund Njeru Njagi, Neal Navani, Diana Sarfati, Bernard Rachet

Comorbidity

Existence of a long-term health condition in the presence of a primary disease of interest Generally, more prevalent with

older age and lower socio-economic position (SEP) Less is known about prevalence of comorbidity in cancer patients Can influence

treatment and outcomes SEP inequalities in cancer treatment and short-term mortality – due to comorbidity?

Aims of study

Use electronic health records to estimate comorbidity prevalence in cancer patients, in order to: Establish prominent comorbid conditions

Identify patterns in comorbidity by SEP (adjusting for age and sex)Four cancers: Colon, rectum, lung, Hodgkin lymphoma (HL)

Data

England National Cancer Registry data Patient / tumour information: Age, sex, SEP (deprivation), type of cancer and date of diagnosis

England hospital admissions records (“Hospital Episode Statistics”, HES) Comorbidity information: Diagnostic information, hospital admission and discharge dates

Fourteen conditions considered as comorbidities

Timeframe for inclusion: Condition diagnosed up to six years prior to cancer diagnosis

Liver disease

Previous malignancyDiabetesObesityDementia

Hemiplegia / paraplegia

Cerebrovascular disease

(

CVD)

Hypertension

Renal disease

Myocardial infarction (MI)

Chronic obstructive pulmonary disease (COPD)

Congestive heart failure (CHF)

Peripheral vascular

disease (PVD)

Rheumatological conditions

Methods of analysis

Prevalence (percentage): crude, and adjusted for age and sexLogistic regression: adjusted* odds ratio of a given condition being present

Multinomial logistic regression: adjusted* probability of condition being present as a single or (part of) multiple comorbidity

*adjusted for age, sex and deprivation

Patients: 15-90 years at diagnosis

Sex

Age

Deprivation groupNo. of patientsColonLung

Rectum

HL

Colon

Rectum

Lung

HL

Patients: 15-90 years at diagnosis

Sex

Age

Deprivation groupNo. of patientsColonLung

Rectum

HL

Colon

Rectum

Lung

HL

Patients: 15-90 years at diagnosis

Sex

Age

Deprivation groupNo. of patientsColonLung

Rectum

HL

Colon

Rectum

Lung

HL

Comorbidity status by cancer type

Percentage of patients with 0, 1 or 2+ comorbidities

Crude and adjusted prevalence (%):

colon cancer

Crude and adjusted prevalence (%):

lung cancer

Crude and adjusted prevalence (%):

HL

Odds ratios of a given condition present,

by deprivation group (Male colon cancer patients, aged 70 years)

Deprivation group(1 = Least deprived, REF)

Odds ratios

of a given condition present, by deprivation group

(Male colon cancer patients, aged 70 years)

Deprivation group(1 = Least deprived, REF)

Probability (%) of having condition as

single or multiple comorbidity(Male colon cancer patients

)

Probability (%) of having condition as single or multiple comorbidity

(Female colon cancer patients)

Conclusions

Common prevalent conditions in each of the four cancersSimilar prevalence patterns by deprivation, age and sex

Deprivation was associated with almost all of 14 comorbid conditions, and with multiple comorbidityPrevalence of some conditions may be underreported in hospital admissions dataImportant insight for investigating the role of specific conditions on cancer outcomes

Acknowledgments

Bernard RachetAurelien BelotLibby EllisCamille Maringe

Miguel Angel Luque-FernandezEdmund Njeru NjagiNeal NavaniDiana SarfatiMichel Coleman

Thank you!