Calculating and using the drug burden index score in research and practice

KEYWORDS:

• Drug burden index
• ageing
• anticholinergic
• sedative
• potentially inappropriate prescribing
• functional outcomes
• deprescribing

1. Introduction

Drug Burden Index (DBI) is a pharmacological measure of an individual’s exposure to medications with anticholinergic and sedative effects [1]. The pharmacological actions of these drugs can reduce physical and cognitive function, particularly in older people with reduced physiological reserve. These effects can be considered type A predictable adverse drug reactions. They are dose-related and are cumulative across drugs from different classes with anticholinergic or sedative actions.

DBI was developed to estimate the risk of functional impairment from medications in older people with the goal of becoming a clinical risk assessment tool. Clinically, in older people, impaired physical and cognitive function have a multitude of causes including drugs, diseases, and ageing itself. The DBI can help clinicians recognize the cumulative exposure of a patient to medicines with anticholinergic and sedative effects, which could contribute to the individual’s functional impairment. The clinician can then weigh these risks against the potential benefits of the medications in the individual, minimizing drug-related functional impairment.

2. Validation of DBI

DBI was first evaluated in a community-dwelling population of older people from the USA participating in the Health Ageing and Body Composition Study. In this cohort, increasing DBI was found to be associated with impaired physical and cognitive function cross-sectionally [1] and with functional decline longitudinally [2]. Over the past decade, DBI has been calculated in populations of older adults from the USA, Canada, Australia, New Zealand, UK,  and Europe [3]. Increasing DBI is consistently associated with impairment of physical function in populations of older adults across a range of measures including short physical performance battery, activities of daily living and instrumental activities of daily living [3,4]. In pharmaco-epidemiologic studies, higher DBI has also been associated with other adverse geriatric outcomes such as falls, fractures, frailty, hospital admission, and mortality [3,4]. In keeping with studies of other measures of exposure to anticholinergic and sedative medicines, the association between DBI and impaired cognition is less consistent and varies with the measure of cognition studied and the population studied [5].

Recently, studies using DBI as a clinical intervention to prompt review of high-risk prescribing have been initiated (e.g. ACTRN12617000765325, www.anzctr.org.au). While clinicians report that they find this information useful [6], there are no published randomized controlled trials powered to determine whether providing a DBI report for patients results in a reduction in DBI and an improvement in clinical outcomes. In addition, a recent interventional study of medication review with a focus on deprescribing anticholinergic and sedative medications used DBI as an outcome measure [7] and found no difference in the proportion of patients in whom DBI was reduced by at least 0.5 between the intervention and control groups.

3. Calculation of DBI

Consistent and accurate calculation of DBI is important for observational studies to allow synthesis of international data, and for future interventional studies and clinical application, to ensure that it is the validated measure of DBI that is applied.

DBI is calculated for an individual by summing the burdens from every anticholinergic or sedative medicine they take regularly. The original methodology [1] can be applied to any formulary to identify medicines with clinically relevant anticholinergic or sedative properties and to determine the minimum licensed/registered dose for each of these medicines. This information is then applied to the person’s drug list to calculate the drug burden from any anticholinergic or sedative drugs, using the equation:

, where E is the pharmacological effect, D is the daily dose taken, and δ is the minimum licensed daily dose, which is used as an estimate of the DR₅₀ (daily dose required to achieve 50% of the maximal effect at steady state).

There has been a discussion in the literature about the estimate of the DR₅₀ being the minimum daily dose licensed in adults for any indication. This estimate was selected because of the requirement for pharmacokinetic-pharmacodynamic data to license a dose. This means that the minimum licensed dose lies somewhere on the dose-response curve. Since most drugs have a licensed dose range that also includes higher doses, the minimum licensed dose is likely to achieve less than 100% of the maximal effect. The decision to use the lowest licensed dose across all indications was pragmatic, due to the variability in whether the anticholinergic or sedative effect is the therapeutic effect for which the drug is licensed, and the limitations of information available on indication in research and practice.

Several alternative estimates of the DR₅₀ have been proposed, including the WHO Defined Daily Dose (DDD) and geriatric doses. The WHO DDD is based on utilization patterns across the whole population rather than pharmacologic data, and is much higher than the minimum registered dose for some drugs but not for others [8]. Geriatric doses from geriatric dosing handbooks are often based on expert opinion and extrapolation in the absence of pharmacologic data, and therefore are not as consistently evidence based as the licensed doses. In old age, the reductions in hepatic and renal clearance and the pharmacodynamic changes have similar impact on most anticholinergic and sedative drugs. Therefore, use of the minimum adult dose appears to be the most accurate and unbiased estimate of DR₅₀ in the ageing population. However, there is potential to evaluate different approaches in different populations.

Using different estimates of DR₅₀ has a moderate effect on the size of the contribution of a drug to a patient’s DBI. Consider the example of a patient who takes 10 mg temazepam each night. Using the minimum licensed dose on the US Food and Drug Administration label of 7.5 mg gives a DBI of 0.57 (10/ (10 + 7.5)); using the minimum licensed dose by the Australian Therapeutic Goods Administration of 10 mg gives a DBI of 0.50 (10/ (10 + 10)); using the WHO DDD of 20 mg gives a DBI of 0.33 (10/ (10 + 20)); and using the minimum dose from the Australian Medicines Handbook Aged Care Companion of 5 mg gives a DBI of 0.67 (10/ (10 + 5)).

In performing the DBI calculation, each drug (ingredient) has a DBI score between zero and one. For patients taking combination tablets containing more than one ingredient, DBI must be calculated separately for each ingredient. For patients who take more than one formulation with the same ingredient, the patient’s total daily dose for that ingredient across all formulations is summed prior to calculating the DBI for that ingredient. This can be complicated when a patient takes the same drug by more than one route, and each route has different bioavailability and a different minimum licensed dose. In this case, it is recommended that the total daily dose is converted to oral equivalents, using the ratio of the minimum licensed doses for the different formulations or the inverse of the oral bioavailability. The DBI for the ingredient is then calculated using the oral minimum licensed dose as the estimate of DR₅₀. This can be demonstrated through the example of a patient who takes morphine slow release 20 mg bd orally and morphine 5 mg 6 hourly subcutaneously. The total equivalent oral daily dose of morphine (oral bioavailability 0.33) that the patient takes is [(20 mg x 2) + (5 mg x 4 × 3)] = 100 mg. The DBI for morphine in this patient is [(equivalent oral daily dose)/(equivalent oral daily dose + minimum licensed oral daily dose) = 100/(100 + 20) = 0.83.

4. Use of DBI in clinical practice

DBI has potential as a clinical risk assessment tool. DBI can help clinicians recognize anticholinergic and sedative medicines that can impair function and prompt them to minimize their patient’s exposure to these medicines. While it is generally not feasible to perform the calculation manually at the bedside, electronic calculators have been developed with reports to inform clinicians of their patient’s DBI and its associated risks [9]. DBI calculators and reports could be integrated into electronic prescribing through computerized decision support systems.

DBI could also be used in clinical audits. Internationally, prescribing indicators for older people guide quality use of medicines at a practitioner or population level. DBI may be a useful measure of high-risk medications for such indicators. Some existing indicators are based on similar principles to DBI. For example, the Scottish Polypharmacy Indicators identify the proportion of older patients with geriatric syndromes who are prescribed multiple drugs with sedating or anticholinergic effects (excluding antiepileptics) [10].

5. Strengths and limitations of DBI

There are strengths and limitations of DBI as a clinical risk assessment tool in terms of scope, evidence base, and translation.

Regarding scope, the DBI is limited to identifying exposure to and risk from cumulative exposure to medications with anticholinergic and sedative effects. It does not consider other medication classes that can cause adverse effects in older adults and does not provide specific clinical guidance. Restriction to anticholinergic and sedative drugs tailors the risk assessment to those drugs where clinicians are most likely to need support. Identifying other high-risk drug classes, such as anticoagulants and hypoglycaemics, is relatively straightforward, while identifying drugs across multiple classes with anticholinergic effects is reported to be challenging clinically [11]. Adverse effects of high-risk drug classes, such as bleeding from anticoagulants and hypoglycemia from insulin, are relatively easy to recognize clinically without a risk assessment tool, while functional impairment from anticholinergic and sedative mediations can be difficult to distinguish from other causes. The DBI measures the risk of impaired function from anticholinergic and sedative effects of drugs and does not attempt to weigh this risk against the current benefit of the medicine for an individual patient. This approach acknowledges the clinical complexity of making therapeutic decisions with patients, which includes consideration of goals of care, multi-morbidity, and function [12].

The  evidence for the association of increasing DBI with impaired function in older people is strong in terms of pharmacological mechanism and consistent international pharmaco-epidemiologic observational studies [3]. However, there are no published randomized controlled trials demonstrating the clinical efficacy and safety of using DBI as a clinical risk assessment tool.

Key translational issues are feasibility, generalizability, and sustainability. For most clinicians, clinical utility will require integration of a DBI calculator and report into computerized decision support systems. Generalizability of DBI is strong. The association of increasing DBI with adverse geriatric outcomes has been demonstrated for older people across different settings (community, residential aged care facilities, hospitals) [3], including those with and without dementia [13]. DBI can be calculated for any older person where accurate information on their medication use including dose is available, using the local formulary. While this is a strength in terms of international applicability, significant work is required to establish the local database of anticholinergic and sedative drugs and their minimum licensed doses. For sustainability, maintenance requires updating the list of anticholinergic and sedative drugs and their minimum licensed doses, whenever the relevant local formulary is updated.

6. Conclusion

There is strong pharmacologic and pharmaco-epidemiologic evidence to support the association of increasing DBI with adverse geriatric outcomes, but to date there is no data on the safety or efficacy of using DBI as a clinical risk assessment tool on clinical outcomes. There is potential to evaluate DBI as a risk assessment tool in research on deprescribing. Translation into practice internationally appears to be feasible through clinical decision support systems and DBI could be applied within existing services for multidisciplinary medication review.

With the high prevalence of polypharmacy and medication-related problems in our ageing, multi-morbid population, it is increasingly important to align treatment to a patient’s therapeutic aims. For older people who aim to optimize their functional independence, minimizing DBI may become an important part of clinical medication review.

Declaration of interest

SN Hilmer developed and continues to conduct an active research program on Drug Burden Index including studies funded by research grants. The author has no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Acknowledgments

SN Hilmer acknowledges the mentorship of Professor Darrell Abernethy for her work on Drug Burden Index. She also acknowledges the contributions of her mentees, particularly Dr Danijela Gnjidic and Dr Lisa Kouladjian-O’Donnell, and of her collaborators, to the research discussed in this editorial.

Additional information

Funding

This paper was not funded.