The effect of multimorbidity on diagnostic interval for lung cancer and mesothelioma: a cohort study using data from the Clinical Practice Research Data link.

Problem

Cancer survival rates are relatively worse in the UK than most other European countries, and this is at least partly due to delays in diagnosis. Understanding the factors contributing to diagnostic delay is crucial to efforts to reduce it. Multimorbidity been suggested to contribute to diagnostic delay via two mechanisms – comorbidities placing “competing demands” on the clinician’s time reducing the chance of possible cancer symptoms being investigated, and “masking conditions’ offering plausible alternative explanations of cancer symptoms. This work investigates the effect of multimorbidity on diagnostic interval for lung cancer and mesothelioma in patients in the UK.

Approach

Patients with incident lung cancer or mesothelioma aged at least 65y in 2019 were identified in the Clinical Practice Research Datalink, those with at least two years of continuous registration between 1990 and 2016 were included. Diagnostic interval was defined as the time from first presentation with a symptom suggestive of lung cancer or mesothelioma to the diagnosis date (symptoms up to 12 months before diagnosis were considered). Co-morbidities considered were coronary heart disease, depression and anxiety, heart failure, hypertension, chronic kidney disease, osteoporosis, dementia, serious mental illness, epilepsy and diabetes, possible masking conditions considered were asthma, chronic obstructive pulmonary disease, chronic fatigue syndrome and a recent (within 2 years) prescription for an angiotensin-converting-enzyme inhibitor. Other factors considered included gender, age at diagnosis, year of diagnosis, background consultation frequency from 24 to 12 months before diagnosis, increase in consultation frequency in the year before diagnosis, presenting symptom and lifestyle variables (smoking status, drinking status, and BMI category). Associations between predictors and diagnostic interval were investigated using linear regression.

Findings

Complete data were available for 10424 lung cancer/mesothelioma patients. In adjusted analyses diagnostic interval was longer among patients with masking conditions, increasing by 27.6 (95%CI 22.9 – 32.4) days and 72.0 (65.6, 78.4) days in patients with one and two or more masking conditions respectively compared to those with none. Number of co-morbidities was not associated with diagnostic delay in adjusted analyses. However, both background consultation frequency and an increased consultation rate in the year before diagnosis were independently positively associated with diagnostic interval, which was 23.0 (17.8 28.3) days higher in those with an increased consultation rate. Diagnostic interval was also increased by 12.5 (5.8, 19.2) days in ever-smokers versus non-smokers, and by 24.4 (14.0, 34.7) days in underweight patients versus those in the normal weight range.

Consequences

The presence of masking conditions offering alternative explanations for lung cancer/mesothelioma symptoms is associated with delayed diagnosis. Patients with higher consultation frequencies also had longer diagnostic intervals, suggesting competing demand may be an issue. Strategies to reduce diagnostic interval in these patients should be considered.