Abstract
Background
Comprehensive medication utilisation and its association with outcomes are not well described in older surgical inpatients.
Objectives
Investigate medication utilisation and the associations of the number of medications and exposure to anticholinergic and sedative medications measured using the Drug Burden Index (DBI exposure) with outcomes in older surgical inpatients.
Methods
A multi-centre retrospective cohort study of surgical patients ≥65 years at five Australian hospitals admitted for ≥48 hours, from 4 January 2022 to 3 January 2023. The number of regular medications was counted on admission and discharge. DBI exposure throughout admission was calculated using the area under the curve for DBI/days of admission (AUC DBI/days). Multilevel regression was conducted for outcomes on/during admission [falls, adverse drug events (ADEs), delirium, pressure ulcers, Hospital Frailty Risk Score (HFRS), length of stay (LOS)].
Results
The 13 815 participants (51.1% male) had a median (IQR) age of 77 (71–84) years. The median number of medications was 5 (2–9) on admission and 9 (6–13) on discharge. The median AUC DBI/days was 0.13 (0–0.5). The number of admission medications was associated with delirium [adjusted odds ratio (aOR) 1.02, 95% CI: 1.01–1.04], ADEs (aOR 1.03, 95% CI: 1.01–1.04), pressure ulcers (aOR 1.04, 95% CI:1.02–1.06), HFRS [adjusted rate ratio (aRR) 1.01, 95% CI:1.001–1.02] and increased LOS (aRR 1.011, 95% CI: 1.008–1.014). The AUC DBI/days was associated with falls (aOR 1.29, 95% CI: 1.18–1.41), delirium (aOR 1.63, 95% CI:1.46–1.82), ADEs (aOR 1.26, 95% CI: 1.11–1.43), pressure ulcers (OR 1.60, 95% CI: 1.38–1.86), HFRS (aRR 1.27, 95% CI:1.17–1.37) and increased LOS (aRR 1.36, 95% CI:1.32–1.40).
Conclusion
Polypharmacy is common amongst older surgical inpatients, and the number of medications increased from admission to discharge. Increased DBI exposure was associated with adverse outcomes.
Keywords: Drug Burden Index, polypharmacy, surgery, older people
Key Points
The majority of older surgical inpatients were exposed to polypharmacy and to anticholinergic or sedative medications.
Medication changes are common; most patients (76.2%) had an increase in the number of medications from admission to discharge.
Exposure to the DBI was associated with delirium, falls, adverse drug events, pressure areas, frailty and length of stay.
Background
Polypharmacy is common in older adults undergoing surgery. A population-based cohort study found that 54.8% of patients ≥65 years undergoing surgery had pre-operative polypharmacy (≥5 medications) [1]. Most studies have investigated exposure to polypharmacy prior to admission. However, new post-operative polypharmacy occurs in 33.4% of adults not exposed to polypharmacy on admission [2]. A recent systematic review of 45 studies found that polypharmacy was associated with postoperative delirium, discharge to a care facility and 30-day mortality in adults undergoing surgery [3]. Polypharmacy was inconsistently associated with other adverse outcomes including prolonged hospitalisation and readmission [3].
Patients undergoing surgery are often taking or are newly commenced on sedative and anticholinergic medications [4, 5], and anaesthetic agents themselves have anticholinergic and sedative effects [6, 7]. These exposures have been associated with adverse outcomes in older surgical patients in some studies. In a large population-based cohort study of older patients undergoing elective noncardiac surgery, anticholinergic drug exposure increased the risk of longer length of stay, readmission, institutional discharge and mortality [8]. While some studies have shown that anticholinergic drug exposure is associated with delirium in older patients requiring a surgical admission [8–11], another study suggested that there was no association between the anticholinergic load and the development of postoperative delirium in older patients without severe preexisting cognitive impairment undergoing elective surgery [12]. Pre-operative benzodiazepine and Z-drug use is common in adults undergoing surgery and is associated with adverse postoperative events, particularly when coprescribed with opioids [13]. Pre-operative opioid use may be associated with prolonged length of stay, worse overall pain, function and quality of life in adults undergoing joint and spine surgery [14, 15]. Additionally, a systematic review found that opioid prescribing at surgical discharge does not reduce pain but increases adverse events [16].
Many anticholinergic and sedative burden scales exist, but most do not account for medication dose [17, 18]. The Drug Burden Index (DBI) is a measure of an individual’s total exposure to both anticholinergic and sedative medications [19]. The DBI is calculated by summing the burden for each regular medication with anticholinergic or sedative effects, based on the dose taken relative to the minimum effective daily dose [19]. The DBI applies the principle of dose response to determine the effect of medication exposure [19]. The DBI has been used as a risk assessment tool in hospital and community settings, and a higher DBI may be associated with adverse clinical outcomes such as falls, impaired function and cognition [20]. Very few studies have analysed DBI exposure dynamically (e.g. area under the curve) and only in community or rehabilitation cohorts [21–23].
While the prevalence of polypharmacy and anticholinergic burden in older surgical patients has been reported [1, 8], no studies have specifically investigated the DBI throughout the admission for surgical patients. Given the frequent medication and dose changes during hospitalisation, and the common indications for anticholinergic and sedative medications peri-operatively, the DBI may be a particularly relevant indicator of high-risk medication exposure in surgical inpatients. Additionally, given that the majority of surgical inpatients are exposed to ≥5 medications, it is important to consider whether receiving a greater number of medications is associated with adverse outcomes, rather than using a specific cut-off. Existing studies on medication changes during surgical admissions have been limited to specific medications or cohorts [24]. Furthermore, although some studies have reported associations of polypharmacy or anticholinergic exposure with delirium, discharge to a care facility or mortality, associations with frailty, pressure ulcers and falls have not been explored in this population.
The objective of this study was to describe medication utilisation and its association with outcomes amongst all older patients admitted under surgical services across five Australian hospitals. The specific aims were to (i) evaluate the distribution of the DBI and number of medications on admission; (ii) evaluate the changes in the DBI and in the number of medications taken during admission; (iii) explore prescribing patterns of DBI contributing medications during admission at the level of service and hospital; (iv) determine the association of admission number of medications and exposure to the DBI throughout admission with clinical outcomes during admission (including length of stay, frailty, falls, adverse drug events, delirium, pressure ulcers and in-hospital mortality); and (v) determine the association between discharge number of medications and the DBI throughout the admission with 28 days re-presentation to hospital.
Methods
Study design
A multi-centre retrospective cohort study using de-identified data was conducted. The study was approved by the institutional ethics committee (2022/ETH01284). A waiver of consent was granted.
Population and setting
Participants were adults aged ≥65 years admitted under surgical teams at five teaching hospitals (Hospital A, B, C, D, E) in New South Wales for ≥48 hours over a 12-month period from 4 January 2022 to 3 January 2023. All admissions under surgical specialties were included. Hospital A was a metropolitan tertiary hospital (~700 beds), Hospitals B and C were metropolitan secondary hospitals (~200 beds), Hospital D was a regional tertiary hospital (~500 beds) and Hospital E was a regional secondary hospital (~300 beds). Surgical services included Orthopaedic Surgery, General Surgery, Urology, Vascular Surgery, Neurosurgery, Acute Surgical Units, Cardiothoracic Surgery, Gastro-Intestinal Surgery, Plastic Surgery, Ear Nose and Throat Surgery, Colorectal Surgery, Hand Surgery, Trauma, Breast Surgery and Faciomaxillary Surgery. Each admission was treated as a separate encounter.
Data collection
De-identified data were extracted from the hospital electronic medical records including regular medications (active ingredients only), medication doses, medication frequency and the DBI. Pro re nata (PRN) medications were excluded. The number of medications on admission and discharge was defined by the number of regular medications charted at each respective time point. The DBI was automatically calculated in the electronic medical record each time regular medications were ordered [25]. The original DBI equation was used; DBI = ∑[D/(δ + D)] where ∑ is the sum score of the prescribed drugs, D is the daily dose taken by the patient, and δ is the minimum daily dose registered for use in adults [26]. DBI exposure during admission was defined as the area under the curve for DBI divided by length of stay (AUC DBI/days). This was calculated by applying the trapezoidal rule using the DBI, time of medication orders and time of admission and discharge, thus accounting for changes in medications and doses throughout admission [19, 22]. A one-unit increase in AUC DBI/days is equivalent to receiving an average of two DBI-contributing medications at the minimum daily dose registered for use in adults throughout the admission.
Study outcomes including falls, adverse drug events, delirium, pressure ulcers and frailty [using the Hospital Frailty Risk Score (HFRS)] [27] were identified using International Classification of Diseases 10th Revision (ICD-10) codes as outlined in Appendix Table S1. It was not possible to determine whether these outcomes were present at admission or developed during hospital admission. Death during admission was determined from discharge disposition data. Length of stay and 28-day re-presentation to the hospital were calculated from admission and discharge times.
Baseline characteristics included age, sex and admitting surgical specialty. The Charlson Comorbidity Index (CCI) [28] was calculated from ICD-10 codes. Emergency admissions were defined by referral from the emergency department, elective by outpatient referral; where unclear, classification was based on clinical judgement by the researcher using admission details. It was not possible to reliably determine whether participants had surgery during admission using ICD codes. A manual check of 100 elective surgery patients found that most participants who had undergone surgery did not have a relevant ICD-10 code.
Analysis
Descriptive statistics were used to describe the demographic and clinical characteristics of the cohort. Violin plots were used to visualise the distribution and changes in medications and DBI during admission. The association between the number of medications and AUC DBI/days with binary outcomes (including falls, adverse drug events, delirium, pressure ulcers, death and 28-day re-presentation to hospital) was analysed using multilevel logistic regression. The association between the number of medications and AUC DBI/days with continuous outcomes (including frailty and length of stay) was analysed using multilevel gamma regression. All regression models were adjusted for confounders including age, sex, comorbidities and length of stay, except when length of stay was analysed as an outcome. Given the variation in DBI-contributing medication prescribing across medical specialties, surgical specialty was included as a random effect [29]. For random effect selection, considering the hierarchical structure of patients within medical services and hospitals, intraclass correlation coefficients (ICCs) were calculated for each binary outcome. Comparison of ICCs between surgical services and hospitals showed that surgical service ICCs exceeded hospital ICCs for all outcomes, with falls showing the highest surgical service ICC of 0.32. Based on these findings, surgical service was selected as the random effect for all multilevel models. In cases where the linearity assumption was violated, the continuous variable was categorised into clinically meaningful groups. Subgroup analysis was performed to compare elective and emergency surgery. Data processing was performed using Python version 3.12. Multilevel analysis was performed using the lme4 package in R version 4.2.0. The code is available on request to the authors. Two-sided P-values <.05 were considered statistically significant.
Results
Over the 12-month period, 13 815 eligible participants were identified. Baseline characteristics are shown in Table 1. The median age of participants was 77.0 years and 51.1% were male. At the time of or during admission, 23.0% participants had falls, 9.3% participants had delirium, 6.3% participants had adverse drug events and 3.7% had pressure ulcers. During admission, 3.3% died and 34.0% re-presented to hospital within 28 days after discharge. Four participants with missing data were excluded from regression analyses.
Table 1.
Baseline characteristics of participants
| Variable | Median (IQR) or n (%) |
|---|---|
| Age | 77.0 (71.0–84.0) |
| Sex | |
| Male | 7054 (51.1) |
| Female | 6761 (48.9) |
| Length of stay | 6.0 (3.4–11.9) |
| Charlson Comorbidity Index | 0 (0–2) |
| Number of medications on admission | 5 (2–9) |
| Area under the curve for Drug Burden Index/length of stay (days) | 0.13 (0–0.55) |
| Hospital | |
| Hospital A | 5455 (39.5) |
| Hospital B | 1662 (12.0) |
| Hospital C | 792 (5.7) |
| Hospital D | 4425 (32.0) |
| Hospital E | 1481 (10.7) |
| Service | |
| Orthopaedic Surgery | 4154 (30.1) |
| General Surgery | 3354 (24.3) |
| Urology | 1199 (8.7) |
| Vascular Surgery | 913 (6.6) |
| Neurosurgery | 879 (6.4) |
| Acute Surgical Unit | 663 (4.8) |
| Cardiothoracic Surgery | 655 (4.7) |
| Gastro-Intestinal Surgery | 612 (4.4) |
| Plastic Surgery | 487 (3.5) |
| Ear Nose & Throat Surgery | 317 (2.3) |
| Colorectal | 285 (2.1) |
| Hand Surgery | 133 (1.0) |
| Trauma | 85 (0.6) |
| Breast Surgery | 61 (0.4) |
| Faciomaxillary Surgery | 18 (0.1) |
| Type of admission | |
| Emergency | 10,701 (77.5) |
| Elective | 3114 (22.5) |
The median (IQR) number of medications on admission was five [2–9] and on discharge was nine [6–13]. The median (IQR) peak number of medications during admission was 12 [8–16]. Compared to admission, 10,538 participants (76.3%) had an increase in the number of medications at discharge, 1432 (10.4%) participants had no change and 1845 (13.4%) had a decrease. The median (IQR) AUC DBI/days was 0.13 (0–0.55). There were 3740 participants (27.1%) who had an increase in the DBI at discharge compared to on admission, 8814 (63.8%) where there was no change and 1261 (9.1%) where there was a decrease. The changes in the number of medications and the DBI between admission and discharge for the cohort are shown in the violin plot in Figure 1. The changes in the number of medications and DBI during admission by hospital and specialty are described in Appendix Tables S2 and S3 and Appendix Figures S1 and S2.
Figure 1.
Comparison of admission number of medications and discharge number of medications and admission Drug Burden Index (DBI) and discharge DBI by hospital.
The most commonly prescribed DBI contributing medications during admission were oxycodone (69.8% participants), cyclizine (29.0% participants), morphine (25.2% participants), tapentadol (17.7% participants) and metoclopramide (14.9% participants). Oxycodone was the most commonly prescribed DBI contributing medication for every surgical specialty. There were some differences depending on the surgical subspecialty and hospital, e.g. levetiracetam was commonly prescribed for neurosurgical patients and pregabalin for vascular surgery patients (Appendix Table S4). The most commonly prescribed non-DBI contributing medications were paracetamol (13 101 participants), enoxaparin (8458 participants), ondansetron (8405 participants), macrogol (7289 participants) and pantoprazole (7010 participants).
Associations between medication exposures and outcomes are shown in Table 2. On multilevel logistic regression, a greater number of admission medications was statistically significantly associated with delirium (OR 1.02, 95% CI: 1.01–1.04), adverse drug events (OR 1.03, 95% CI: 1.01–1.04) and pressure ulcers (OR 1.04, 95% CI: 1.02–1.06). There was no significant association with falls. Compared to those who had 0–1 medication on admission, there was a lower risk of death in those on 2–4 medications (OR 0.68, 95% CI: 0.51–0.91), 5–9 medications (OR 0.69, 95% CI: 0.53–0.88) and 10+ medications (OR 0.64, 95% CI: 0.48–0.84). A higher AUC DBI/days was associated with falls (OR 1.29, 95% CI: 1.18–1.41), delirium (OR 1.63, 95% CI: 1.46–1.82), adverse drug events (OR 1.26, 95% CI: 1.11–1.43) and pressure ulcers (OR 1.60, 95% CI: 1.38–1.86). There was no significant association with mortality. On gamma regression, a higher number of medications on admission was associated with frailty (RR 1.01, 95% CI: 1.001–1.02). A higher AUC DBI/days was also associated with frailty (RR 1.27, 95% CI: 1.17–1.37). A greater number of medications (RR 1.011, 95% CI: 1.008–1.014) and AUC DBI/days (RR 1.36, 95% CI: 1.32–1.40) were associated with increased length of stay. There was no association between the discharge number of medications (OR 0.997, 95% CI: 0.988–1.005) or discharge DBI (OR 1.07, 95% CI: 0.998–1.14) and 28-day re-presentation to hospital. Subgroup analyses for emergency and elective admissions are presented in Appendix Tables S5 and S6. The association between medication exposure and 28-day re-presentation to hospital, as well as subgroup analyses, is shown in Appendix Table S7. The association between a lower number of medications on admission and death for emergency admissions was similar to the whole cohort, while there was no association between the number of medications and death for elective admissions.
Table 2.
Association of number of admission medications or area under the curve for DBI/days (AUC DBI/days) with outcomes during admission
| Outcome | n (%) or median (IQR) | Number of medications on admission [OR or RR (95% CI)] | P-value | AUC DBI/days [OR or RR (95% CI)] | P-value |
|---|---|---|---|---|---|
| Falla | 3179 (23.0) | OR 0.99 (0.98–1.00) | .052 | OR 1.29 (1.18- 1.41) | <.001 |
| Deliriuma | 1282 (9.3) | OR 1.02 (1.01–1.04) | <.001 | OR 1.63 (1.46–1.82) | <.001 |
| Adverse drug eventsa | 865 (6.3) | OR 1.03 (1.01–1.04) | <.001 | OR 1.26 (1.11–1.43) | <.001 |
| Pressure ulcera | 507 (3.7) | OR 1.04 (1.02–1.06) | <.001 | OR 1.60 (1.38–1.86) | <.001 |
| In hospital mortalitya | 455 (3.3) | 0–1 medication: 1.0 (reference) 2–4 medications: OR 0.68 (0.51–0.91) 5–9 medications: OR 0.69 (0.53–0.88) 10+ medications: OR 0.64 (0.48–0.84) |
– .009 .004 .001 |
OR 1.18 (0.97–1.43) | .10 |
| Frailty (HFRS)a | 3.1 (0.6–6.9) | RR 1.01 (1.001–1.02) | .02 | RR 1.27 (1.17–1.37) | <.001 |
| Length of stayb | 6.0 days (3.4–11.9 days) | RR 1.011 (1.008–1.014) | <.001 | RR 1.36 (1.32–1.40) | <.001 |
OR, odds ratio; RR, rate ratio; CI, confidence interval; IQR, interquartile range; DBI, Drug Burden Index; AUC, area under the curve; HFRS, Hospital Frailty Risk Score. All analyses included surgical specialty as a random effect.
aAdjusted for age, sex, comorbidity and length of stay
bAdjusted for age, sex and comorbidity
Discussion
This study found that polypharmacy, anticholinergic and sedative medication use and medication changes are common in older adults admitted under surgical teams. The median number of medications participants were receiving on admission was five. On average, participants were prescribed four additional medications at discharge and 76.3% of participants had an increase in the number of medications at discharge compared to admission. Anticholinergic and sedative medication exposure, measured using AUC DBI/days, which considers the type and doses of medications, as well as changes during hospitalisation, was associated with adverse outcomes. For example, for every additional admission medication, the odds of delirium increased by 2%; while for every one unit increase in AUC DBI/days, the odds of delirium increased by 63%.
The prevalence of polypharmacy in this study was similar to previous studies in older inpatients. Like a previous study conducted in Iceland, which included adults undergoing surgery, new polypharmacy was common during hospitalisation [2]. Our study investigated the prevalence of medication changes during surgical admissions, which has not been comprehensively explored previously. Our findings on the prevalence of anticholinergic and sedative medication use were the first to consider total exposure including dose but were consistent with previous studies using different measures of anticholinergic burden or single drug classes with these effects [8, 12]. Opioids were commonly prescribed in this population with 69.8% of participants receiving regular oxycodone at some stage during their admission. This is similar to the prevalence of opioid use in opioid-naïve adults undergoing low-risk surgeries reported in a previous study [30] although higher than the prevalence of opioid use on discharge after major surgery reported in another study [31]. Given the prevalence and risks of polypharmacy and opioid use in this population as well as the frequent changes in medication during hospitalisation, this is a population at risk of medication-related harms and proactive medication review is important.
The number of medications on admission was associated with some adverse outcomes including delirium, adverse drug events, pressure ulcers, frailty and increased length of stay. However, the effect size was modest. Previous studies analysing the association between polypharmacy and outcomes in surgical inpatients have used numerical definitions of polypharmacy, such as five or more medications [3]. Similar to previous studies, this study showed an association between polypharmacy and increased length of stay and delirium in surgical inpatients. While the association between polypharmacy and pressure ulcers, falls and frailty has been shown in other populations [32–34], it has not been widely explored in those undergoing surgery.
There was a counterintuitive association between fewer medications on admission and adverse outcomes including mortality and falls for emergency admissions. This contrasts with previous studies that have suggested that polypharmacy is associated with an increased risk of death in those undergoing surgery [3]. This difference may be due to clinical quality issues surrounding admission medications or a clinical decision to withhold nonessential medications due to acuity of illness. Medication omissions are common in patients who are acutely unwell in the hospital setting [35]. Those who were more acutely unwell may have had a single regular medication charted rather than all usual medications. Further studies, linking to dispensed medication databases to determine medications prior to admission, may be useful to determine whether the number of medications prior to admission impacts death or falls.
This is the first study to investigate the association of the DBI with outcomes in surgical inpatients, although previous studies have investigated the associations of exposure to anticholinergic medications and outcomes in this population [8–12]. We found a strong association between exposure to medications with anticholinergic and sedative effects throughout the admission (calculated using the AUC DBI/days) with adverse clinical outcomes. Similar to previous cohort studies looking at anticholinergic drug use, we found that high-risk medication exposure was associated with increased length of stay and delirium [8–11]. However, we did not observe an association with mortality during admission or 28-day re-presentation to hospital.
A strength of the study was that it was a large multi-site study. However, it has several limitations. As participants were from one state in Australia, findings may not be generalisable to other populations. The data extract only contained data from during the participant’s admission. The number of medications on admission was calculated based on the number of regular medications charted. This assumes that usual medications were charted correctly initially. In practice, medication errors on admission, particularly medication omissions, are common [36]. PRN medications were not included, and it is likely that participants also received PRN analgesics and anti-emetics, amongst other medications, underestimating their total medication exposure. There was also the assumption that charted medications were administered. Changes to medication orders may indicate intentional prescribing changes, correction of prescribing errors on admission, or new prescribing errors during the admission. In determining some outcomes, ICD-10 codes were used. There is a risk of coding errors and underreporting. Additionally, it was not possible to distinguish whether the outcome occurred at the time of or during the admission. It is likely that some outcomes, e.g. falls in patients admitted under Orthopaedics, were the reason for presentation. There are limitations of cross-sectional retrospective outcome data and a causal relationship cannot be established. Additionally, it was not possible to adjust for some confounding factors, such as if patients underwent surgery. In determining whether patients were elective or emergency patients, it is possible that some elective participants were advised to present through the emergency department and were misclassified as emergency admissions.
The DBI calculator can be integrated into electronic medical records with a computerised clinical decision support system to enable use in practice [25]. Further studies examining whether there is a causal relationship between medication exposure and harms are required. Additionally, it is unclear if interventions to reduce polypharmacy and deprescribe high-risk medications could impact adverse outcomes in this setting [3]. Comprehensive geriatric assessment, which includes a medication review, has been shown to improve outcomes in patients undergoing surgery [37]. While there have been studies on the impact of deprescribing during the pre-operative period for older adults undergoing surgery, there have been no studies analysing the impact of deprescribing interventions on the DBI in older surgical inpatients [3]. There are likely to be additional challenges in implementing deprescribing interventions in the inpatient surgical setting as knowledge of deprescribing is limited and doctors working in this setting are unlikely to proactively deprescribe medications [38]. Future studies investigating the impact of medication review and deprescribing in older surgical inpatients are required.
Polypharmacy and high-risk prescribing are common in older adults admitted to surgical services. In this study, the number of medications and exposure to anticholinergic and sedative medications (AUC DBI/days) were associated with some adverse outcomes with varying effect sizes across different outcomes and exposure measures. The AUC DBI/days was a determinant of adverse outcomes. Further studies, linking with community data from dispensed medication databases and adjusting for additional confounders, such as the acuity of presentation and type of surgery, may provide more accurate insights. The impact of deprescribing interventions on medication-related and clinical outcomes should be further explored in older surgical inpatients.
Supplementary Material
Contributor Information
Bonnie Mengyuan Liu, Ageing and Pharmacology Laboratory, Kolling Institute, Faculty of Medicine and Health, The University of Sydney, Sydney, New South Wales, Australia; Aged Care Department, Royal North Shore Hospital, St Leonards, New South Wales, Australia.
Kenji Fujita, Ageing and Pharmacology Laboratory, Kolling Institute, Faculty of Medicine and Health, The University of Sydney, Sydney, New South Wales, Australia.
Janani Thillainadesan, Department of Geriatric Medicine and Centre for Education and Research on Ageing, Concord Repatriation General Hospital, Concord, New South Wales 2139, Australia.
Danijela Gnjidic, Sydney Pharmacy School, Faculty of Medicine and Health, University of Sydney, Sydney, New South Wales, Australia.
Sarah N Hilmer, Ageing and Pharmacology Laboratory, Kolling Institute, Faculty of Medicine and Health, The University of Sydney, Sydney, New South Wales, Australia; Clinical Pharmacology Department, Royal North Shore Hospital, St Leonards, New South Wales 2065, Australia.
Declaration of Conflicts of Interest
S.H. developed and continues to lead an active research program on the Drug Burden Index. The Goal-directed Medication review Electronic Decision Support System (G-MEDSS) ©, which includes a Drug Burden Index Calculator ©, is under consideration for commercialisation involving S.H. and D.G.
Declaration of Sources of Funding
This study was supported by the Australian National Health and Medical Research Council (NHMRC) Targeted Call for Research into Frailty in Hospital Care (APP 1174447) and the NHMRC 2022 Centres of Research Excellence grant ‘Frailty ADD: Improving Hospital Outcomes for Frail Patients Across Different Disciplines’ (APP 2015821). The funder had no role in the design, execution, analysis and interpretation of data or writing of the study.
References
- 1. McIsaac DI, Wong CA, Bryson GL et al. Association of Polypharmacy with survival, complications, and healthcare resource use after elective noncardiac surgery: a population-based cohort study. Anesthesiology 2018;128:1140–50. 10.1097/ALN.0000000000002124. [DOI] [PubMed] [Google Scholar]
- 2. Jónsdóttir F, Blöndal AB, Guðmundsson A et al. Epidemiology and association with outcomes of polypharmacy in patients undergoing surgery: retrospective, population-based cohort study. BJS Open 2023;7:zrad041. 10.1093/bjsopen/zrad041. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Lee Y, Liew Y, Sim MH et al. Association of polypharmacy and perioperative outcomes: systematic review and meta-analysis. J Perioper Pract 2025;0. 10.1177/17504589251320818. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Young JC, Dasgupta N, Chidgey BA et al. Postsurgical opioid prescriptions and risk of long-term use: an observational cohort study across the United States. Ann Surg 2021;273:743–50. 10.1097/SLA.0000000000003549. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Evans T, Binns H, Mandal AK et al. The impact of anticholinergic burden on clinical outcomes in older hospitalised surgical patients. Br J Hosp Med (Lond) 2024;85:1–9. 10.12968/hmed.2023.0034. [DOI] [PubMed] [Google Scholar]
- 6. Brown KE, Mirrakhimov AE, Yeddula K et al. Propofol and the risk of delirium: exploring the anticholinergic properties of propofol. Med Hypotheses 2013;81:536–9. 10.1016/j.mehy.2013.06.027. [DOI] [PubMed] [Google Scholar]
- 7. Coleman R, Zeinali D, Hall A. Side effects of general anaesthetics and therapeutic gases. Side Eff Drugs Annu 2023; 45:115–35. 10.1016/bs.seda.2023.08.002. [DOI] [Google Scholar]
- 8. McIsaac DI, Wong CA, Diep D et al. Association of Preoperative Anticholinergic Medication Exposure with postoperative healthcare resource use and outcomes: a population-based cohort study. Ann Surg 2019;270:1049–57. 10.1097/SLA.0000000000002765. [DOI] [PubMed] [Google Scholar]
- 9. Mohanty S, Lindroth H, Timsina L et al. A mediation analysis examining high risk, anticholinergic medication use, delirium, and dementia after major surgery. J Surg Res 2024;298:222–9. 10.1016/j.jss.2024.03.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Ieong C, Chen T, Chen S et al. Differences of anticholinergic drug burden between older hospitalized patients with and without delirium: a systematic review and meta-analysis based on prospective cohort studies. BMC Geriatr 2024;24:599. 10.1186/s12877-024-05197-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Herrmann ML, Boden C, Maurer C et al. Anticholinergic drug exposure increases the risk of delirium in older patients undergoing elective surgery. Front Med (Lausanne) 2022; 9:871229. 10.3389/fmed.2022.871229. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Heinrich M, Müller A, Cvijan A et al. Preoperative comparison of three anticholinergic drug scales in older adult patients and development of postoperative delirium: a prospective observational study. Drugs Aging 2021;38:347–54. 10.1007/s40266-021-00839-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Gaulton TG, Wunsch H, Gaskins LJ et al. Preoperative sedative-hypnotic medication use and adverse postoperative outcomes. Ann Surg 2021;274:e108–14. 10.1097/SLA.0000000000003556. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Line B, Bess S, Gum JL et al. Opioid use prior to surgery is associated with worse preoperative and postoperative patient reported quality of life and decreased surgical cost effectiveness for symptomatic adult spine deformity; a matched cohort analysis. N Am Spine Soc J 2021;9:100096. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Goplen CM, Verbeek W, Kang SH et al. Preoperative opioid use is associated with worse patient outcomes after Total joint arthroplasty: a systematic review and meta-analysis. BMC Musculoskelet Disord 2019;20:234. 10.1186/s12891-019-2619-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Fiore JF Jr, El-Kefraoui C, Chay MA et al. Opioid versus opioid-free analgesia after surgical discharge: a systematic review and meta-analysis of randomised trials. Lancet 2022;399:2280–93. 10.1016/S0140-6736(22)00582-7. [DOI] [PubMed] [Google Scholar]
- 17. Hilmer SN, Gnjidic D. The anticholinergic burden: from research to practice. Aust Prescr 2022;45:118–20. 10.18773/austprescr.2022.031. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Braithwaite E, Todd OM, Atkin A et al. Interventions for reducing anticholinergic medication burden in older adults-a systematic review and meta-analysis. Age Ageing 2023;52:afad176. 10.1093/ageing/afad176. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Hilmer SN, Mager DE, Simonsick EM et al. A drug burden index to define the functional burden of medications in older people. Arch Intern Med 2007;167:781–7. 10.1001/archinte.167.8.781. [DOI] [PubMed] [Google Scholar]
- 20. Liu BM, Kouladjian O'Donnell L, Redston MR et al. Association of the Drug Burden Index (DBI) exposure with outcomes: a systematic review. J Am Geriatr Soc 2024;72:589–603. 10.1111/jgs.18691. [DOI] [PubMed] [Google Scholar]
- 21. Ogawa Y, Nibe F, Ogawa R et al. Anticholinergic and sedative drug burden and functional recovery after cerebrovascular accident: a retrospective descriptive study. Prog Rehabil Med 2020;5:20200010. 10.2490/prm.20200010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Hilmer SN, Mager DE, Simonsick EM et al. Drug burden index score and functional decline in older people. Am J Med 2009;122:1142–1149.e2. 10.1016/j.amjmed.2009.02.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Wilson NM, Hilmer SN, March LM et al. Associations between drug burden index and physical function in older people in residential aged care facilities. Age Ageing 2010;39:503–7. 10.1093/ageing/afq053. [DOI] [PubMed] [Google Scholar]
- 24. Sharma NK, Olotu B, Mathew A et al. Lumbar spine surgeries and medication usage during hospital stay: one-Center perspective. Hosp Pharm 2017;52:774–80. 10.1177/0018578717737431. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Kouladjian L, Gnjidic D, Chen TF et al. Development, validation and evaluation of an electronic pharmacological tool: the drug burden index calculator©. Res Social Adm Pharm 2016;12:865–75. 10.1016/j.sapharm.2015.11.002. [DOI] [PubMed] [Google Scholar]
- 26. Hilmer SN. Calculating and using the drug burden index score in research and practice. Expert Rev Clin Pharmacol 2018;11:1053–5. 10.1080/17512433.2018.1528145. [DOI] [PubMed] [Google Scholar]
- 27. Gilbert T, Neuburger J, Kraindler J et al. Development and validation of a hospital frailty risk score focusing on older people in acute care settings using electronic hospital records: an observational study. Lancet 2018;391:1775–82. 10.1016/S0140-6736(18)30668-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Charlson ME, Pompei P, Ales KL et al. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40:373–83. 10.1016/0021-9681(87)90171-8. [DOI] [PubMed] [Google Scholar]
- 29. Kiang MV, Humphreys K, Cullen MR et al. Opioid prescribing patterns among medical providers in the United States, 2003-17: retrospective, observational study. BMJ 2020;368:l6968. 10.1136/bmj.l6968. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Ladha KS, Neuman MD, Broms G et al. Opioid prescribing after surgery in the United States, Canada, and Sweden. JAMA Netw Open 2019;2:e1910734. 10.1001/jamanetworkopen.2019.10734. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Clarke H, Soneji N, Ko DT et al. Rates and risk factors for prolonged opioid use after major surgery: population based cohort study. BMJ 2014;348:g1251. 10.1136/bmj.g1251. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Park-Lee E, Caffrey C. Pressure ulcers among nursing home residents: United States, 2004. NCHS Data Brief 2009;14:1–8. [PubMed] [Google Scholar]
- 33. Ming Y, Zecevic A. Medications & Polypharmacy Influence on recurrent fallers in community: a systematic review. Can Geriatr J 2018;21:14–25. 10.5770/cgj.21.268. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Ekram ARMS, Woods RL, Ryan J et al. The association between polypharmacy, frailty and disability-free survival in community-dwelling healthy older individuals. Arch Gerontol Geriatr 2022;101:104694. 10.1016/j.archger.2022.104694. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Warne S, Endacott R, Ryan H et al. Non-therapeutic omission of medications in acutely ill patients. Nurs Crit Care 2010;15:112–7. 10.1111/j.1478-5153.2010.00365.x. [DOI] [PubMed] [Google Scholar]
- 36. Roughead EE, Semple SJ, Rosenfeld E. The extent of medication errors and adverse drug reactions throughout the patient journey in acute care in Australia. Int J Evid Based Healthc 2016;14:113–22. 10.1097/XEB.0000000000000075. [DOI] [PubMed] [Google Scholar]
- 37. Thillainadesan J, Hilmer SN, Fleury AM et al. New horizons in the perioperative care of older adults. Age Ageing 2022;51:afab245. 10.1093/ageing/afab245. [DOI] [PubMed] [Google Scholar]
- 38. Liu B, Thillainadesan J, Langford A et al. Patient, carer and healthcare professional perspectives on deprescribing in surgical wards: a mixed methods study. Br J Clin Pharmacol. 2025;91:2684–2695. 10.1002/bcp.70088. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.