Improving the prognosis of dizzy elderly: a three-arm validation and intervention study in general practice

Dizziness is one of the geriatric giants. Thirty percent of people above 65 years of age experience some form of dizziness, increasing to 50% in persons above 85. Dizziness can lead to persistent limitations in daily functioning. It is a major risk factor for falling, leading to further limitation of activities, fractures, and high healthcare costs. With the ageing of the population, the burden of dizziness – on society, health care systems, and individuals – will increase significantly. When treating older patients with dizziness, the current diagnosis-oriented approach – as advocated in clinical guidelines – is inadequate. Often, it is not possible to diagnose an underlying disease, and when a disease is diagnosed, therapeutic options are limited. Therefore, the approach of dizziness in older patients has to change. Physicians should not just stick to their diagnostic quest, but also pay attention to the prognosis of a dizzy patient: is this patient at risk of an unfavourable outcome and, if so, how to improve this outcome? Recently, our research group completed a large cohort study investigating the prognosis of dizzy older patients in general practice. We identified seven risk factors for chronic dizziness with persistent impairment, that is (1) onset of dizziness more than six months ago, (2) dizziness provoked by standing still, (3) trouble with walking and/or falling as associated symptom, (4) polypharmacy, (5) absence of diabetes, (6) anxiety and/or depressive disorder, and (7) impaired functional mobility. Three of these risk factors are amenable to treatment: polypharmacy, anxiety and/or depressive disorder, and impaired functional mobility. The results of this cohort study, combined with previous findings of our research group, provide the basis of the present research proposal. The first objective of this project is to validate a previously developed risk score for chronic dizziness with persistent impairment in a new population of dizzy older patients in general practice. The second objective of this project is to investigate the effectiveness of a risk factor-guided intervention for treating dizzy older patients with a high risk of persistent impairment. We will use a three-arm study design to perform both validation and intervention study. In the validation study, we will externally validate a 7-item risk score (risk factors: see paragraph above) in a new population of patients aged 65 years or older presenting with dizziness in general practice. In the intervention study, we will perform a randomized controlled clinical trial among dizzy older patients in general practice, comparing the effectiveness of a targeted intervention – including medication adjustment, psychotherapy, and physical exercise - with usual care during one year of follow-up. Randomisation will take place on patient level. Eligible patients will be identified by participating general practitioners (GPs) and by means of monthly database searches. Patients will be included if they have consulted their GP for a new episode of dizziness with current dizziness-related impairment. Exclusion criteria are severe cognitive impairment, serious comorbid conditions, or current enrolment in an interfering study. The primary outcome measure for both validation and intervention study is dizziness-related impairment (assessed by the Dizziness Handicap Inventory [DHI]). Secondary outcome measures for the intervention study include anxiety or depression (PRIME-MD Patient Health Questionnaire), dizziness frequency (calendar), fall incidence (calendar), and health care utilization (number of registered medical consultations, prescriptions, referrals, hospital admissions, and nursing home admissions, as extracted from the electronic medical databases). To investigate the external validity of our risk score, we will assess reliability, calibration, and discrimination of the model, and we will adjust for overoptimism by means of bootstrapping. To investigate the effectiveness of our intervention, we will use a repeated-measures mixed-effect (RMME) model to assess differences in the 1-year DHI score change. If our risk score provides accurate predictions, GPs will have an easy-to-use clinical tool to identify dizzy older patients with a poor functional prognosis. If our intervention is effective, GPs will be better able to treat dizzy older patients, even when a diagnosis is (still) not available. This will not only improve patients’ perceived health, but also will reduce healthcare utilization and healthcare costs. On a higher level, our project may contribute to a necessary shift of paradigm in general practice: a prognosis-oriented approach (identifying and treating patients at risk) should not just be reserved for patients with a clear-cut diagnosis, but also for patients with chronic symptoms of unknown origin.