Recording patient co-morbidities: where’s the best place to get the information?

Matt Hale is a National Institute for Health Research funded Academic Clinical Fellow & Speciality registrar in Geriatrics in Yorkshire & Humber. His research interests are in improving appropriate treatment stratification for older adults with frailty with a particular focus on statin therapy.

Patients’ self-reported information on their health conditions is often used in clinical practice. It may be used for direct patient care, when patients are asked about their current medical conditions when admitted acutely to hospital or if they move to a new general practice. Alternatively, self-reported data may be used for research studies when either general practice or hospital data is not available or when these data sources are incomplete. Although some studies have looked at the extent of agreement between individuals’ self-report of their health conditions and their general practice’s record of their health conditions, few have focussed on the community-dwelling older adult population, despite this population being among those who have the most contact with secondary care. Furthermore, only a small number of previous studies included individuals with cognitive impairment and none have investigated the impact of frailty.

As treatment decisions and research findings will often be affected by the accuracy of patients’ self-reported medical conditions, it is imperative to know if we have the correct information. To further investigate this, we studied the agreement between self-reported and general practice recorded data among community-dwelling older adults as well as the impact of participant factors, including frailty, on this agreement.

We studied 964 participants in the Community Ageing Research 75+ (CARE75+) study comparing participants’ self-reported health conditions to the conditions reported on their general practice electronic health records. We assessed for agreement for the following health conditions: any cancer (excluding non-melanoma skin cancer), asthma, cerebrovascular disease, chronic obstructive pulmonary disease, dementia, diabetes mellitus, heart failure, peripheral vascular disease, registered blind or partially sighted and rheumatoid arthritis.

We found a marked variation in agreement between the participants’ self-report and the GP record for health conditions ranging from only fair agreement for rheumatoid arthritis, where many patients reported having this but few had a recorded diagnosis on their GP record, to perfect agreement for asthma and peripheral vascular disease. The extent of agreement was also reduced by the presence of frailty or cognitive impairment. Interestingly the presence of multi-morbidity improved agreement for some health conditions (dementia and being partially sighted) whilst reducing the agreement for others (cancer and diabetes).

In an age of information overload where clinicians and researchers alike have choices as to how to obtain information on patient co-morbidities, it is important to recognise that there is not only a difference between what patients and their GPs report but also that the extent of this disagreement is different with different health conditions and is also affected by multi-morbidity, a person’s frailty and a person’s cognitive state. This study highlights the importance of utilising not only one source of information for finding out older adults’ comorbidities and the importance of ensuring that electronic health records are accurate and up to date.

Read the Age and Ageing paper 'A cross-sectional study assessing agreement between self-reported and general practice-recorded health conditions among community dwelling older adults'