One in six admitted patients with low back pain received high daily opioid dosage in Australian hospitals: a cross-sectional study Authors: Qiuzhe Chen, Christina Abdel Shaheed, Chung-Wei Christine Lin, Chris G Maher, Gustavo C Machado Published: 2026 ABSTRACT AimTo examine the amount of opioids administered to patients with low back pain in the hospital setting and to determine factors associated with high daily dosage.MethodsThis cross-sectional study analysed medical records of three public hospitals in Sydney, NSW, Australia. We included adult patients who presented to the emergency department (ED) from January 2016 to October 2021 with non-specific low back pain or radiculopathy. The opioid dose administered in the ED and the daily dosage during hospital stay (if admitted) were estimated using opioid orders measured by milligram of morphine equivalents (MME). High-dose opioid therapy was defined as ≥ 50 MME/day during hospital stay.ResultsOf the 9,345 eligible ED encounters, we included 12,850 opioid orders placed for 5,132 patients (54.9%). Among the 7,633 patients discharged home from ED, 3,519 patients (46.1%) were administered opioids in the ED at an average dose of 13.9 MME (SD = 35.6). 94.2% of patients who were admitted to hospitals (1,613 out of 1,712) received a mean dosage of 34.7 MME/day (SD = 68.3), including 154 patients (9.0%) received 50–90 MME/day and 117 (6.8%) received ≥ 90 MME/day. High-dose opioid therapy was associated with younger age (aOR = 0.80, 95% CI 0.74–0.85), male gender (aOR = 1.38, 95% CI 1.05–1.81), disadvantaged socio-economic status (aOR = 1.50, 95% CI 1.07–2.10), and serious pathology diagnosed during hospital admission (aOR = 1.79, 95% CI 1.36–2.37).ConclusionAlmost one in six admitted patients with low back pain receive a high daily opioid dosage in Australian hospitals. Strategies implementing evidence-based non-opioid alternatives to reduce high-dose opioid therapy are warranted.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12891-026-09882-y. Introduction Low back pain is one of the most common reasons for patients presenting to emergency departments (EDs) and subsequently being admitted to hospital. In 2024–25, almost 160,000 ED visits in Australia were for low back pain and 30% of patients were admitted with a mean cost of AUD$15,000 [1, 2]. The use of pain medications for low back pain in hospitals does not always align with guideline recommendations. Clinical practice guidelines [3] recommend non-opioid analgesics, such as non-steroidal anti-inflammatory drugs (NSAIDs), as first-line medications for acute low back pain and, for inpatients, opioids should be restricted to those with severe pain not responding to non-opioid therapy [4]. However, previous studies found that around 70% of patients with low back pain presenting to the ED in Australia were administered an opioid; [5, 6] and almost 80% of admitted patients with low back pain, and no serious pathology such as fracture or infection, received opioid therapy during hospital admission [7]. The harm in using opioids for low back pain may outweigh the benefits, with greater risks of serious opioid-related adverse events and opioid-related mortality in older adults [8]. Previous systematic reviews have shown that opioids provide small effects on reducing chronic low back pain compared to placebo and increase the risk of adverse events [9, 10]. The recent OPAL randomised trial found no benefit in acute spinal pain relief using oxycodone/ naloxone over placebo, while more adverse events were reported in the opioid group [11]. After hospital discharge, approximately 7% of patients with low back pain prescribed opioids in the emergency department continued use for six months or longer [12], whereas this proportion increased to one-quarter among patients initiated on opioid therapy during inpatient rehabilitation [13]. Despite guidelines recommending that opioid therapy should only be used at the lowest effective dose, there is a dearth of evidence on the opioid dose prescribed to patients with low back pain, particularly in the hospital setting. High-dose opioid therapy above 50 milligram of morphine equivalents per day (MME/day) has been found to provide less than clinically important treatment effects and progressively increase the risk of adverse events and prolonged opioid use [3, 14]. Therefore, this study aimed to examine the opioid dose administered to patients with low back pain during an ED visit and the daily opioid dosage provided to patients admitted to hospitals via ED and to determine factors associated with high daily dosage. Methods This study is reported in accordance with the Reporting of Studies Conducted using Observational Routinely-collected Data (RECORD) guideline [15]. Setting and participants This cross-sectional study analysed data extracted from the electronic medical records of three major public hospitals in the Sydney Local Health District, NSW, Australia: Royal Prince Alfred Hospital, Concord Repatriation General Hospital, and Canterbury Hospital. The Sydney Local Health District Human Research Ethics Committee (Royal Prince Alfred Hospital zone) granted approval for this study (No. X17-0419). We included patients aged ≥ 18 years who presented to the ED of any of the three hospitals from 1 January 2016 to 31 October 2021 and used SNOMED-CT-AU diagnostic codes at ED discharge to identify patients who were diagnosed with non-serious low back pain in the ED (i.e., non-specific low back pain or radiculopathy). Patients with an ED diagnosis of any serious pathology or aged < 18 years were excluded. ED encounters at Concord Repatriation General Hospital and Canterbury Hospital before 1 November 2017 and 1 April 2019 respectively had no digital records of medication orders and were therefore excluded. Data collection Demographic characteristics of patients included age, gender, residential postcode, country of birth, preferred language of communication, and use of interpreter service. Patients aged 65 years or above were grouped as older adults, as recommended in clinical guidelines [3]. Socioeconomic status was estimated using the Index of Relative Socio-economic Advantage and Disadvantage (IRSAD) by residential suburb [16], with the first to fifth deciles classified as relatively disadvantaged. Patients born in any country outside of Australia were designated as born overseas, and all languages other than English were collapsed into a single “non-English” category. For features of an ED visit, we extracted data on the mode of transport to the ED, time and date of arrival, triage category, ED length of stay, and diagnosis at ED discharge (SNOMED-CT-AU codes). Any modes of transport other than ambulance were considered self-presented. Patients who arrived at the ED between 8:00AM–5:00PM Monday–Friday were classified as visits within working hours. Category 1 and 2 of the Australasian Triage Scale (ATS) [17] represented urgent cases. ED length of stay was timed from triaging to ED discharge, and the cut-off for a prolonged stay was four hours [18]. For admitted patients, we collected their principal diagnosis at inpatient discharge (ICD-10-AM codes) and length of hospital admission. Some admitted patients provisionally diagnosed in the ED with non-serious low back pain were subsequently diagnosed with serious pathology during hospital admission, and thus we grouped inpatient diagnoses into five categories: non-specific low back pain, radiculopathy, serious spinal pathology, serious pathology beyond lumbar spine, and non-serious musculoskeletal conditions beyond lumbar spine [19]. Length of hospital admission was measured by days from ED arrival to inpatient discharge, and we used the median length of admission of included patients, i.e., six days, to determine a prolonged admission. Outcome measures The amount of opioid use for each ED encounter was estimated using opioid orders placed by clinicians. We assumed that all ordered opioids were administered to patients during their stay in the ED and/or inpatient admission. Records of an opioid order included name of the active opioid ingredient, strength, the number of doses, route, form of drug, and time and date of order. The MME conversion factor for each type of active opioid ingredient and the assumptions made for data cleaning and MME conversion calculation are tabulated in the Appendix. MME conversion factors were extracted from the United States Centers for Disease Control and Prevention (CDC) guideline for prescribing opioids 2022 [3] and other sources [20, 21]. Opioid orders were transformed to doses by milligram and then converted to MME. If the opioid strength was not recorded in the order, we made an estimate using the most used strength in clinical practice; otherwise, orders were excluded if the type of opioid ingredient had multiple strengths equally used in practice. Opioid orders measured by units not convertible to MME were also excluded, such as sachet, patch, or millilitre. Statistical analysis Multiple ED encounters for low back pain from the same patient during the study period were independently analysed. All analyses were conducted using RStudio (Version 1.4). We calculated individual total MME opioid dosage by adding up all opioid orders placed for each ED encounter. For admitted patients, as we were unable to identify the precise date and time when opioids were administered, we estimated the daily opioid dosage by dividing the total amount of opioids by the length of hospital stay measured from ED arrival to inpatient discharge. Based on the United States CDC guideline [3], we stratified the mean daily opioid dosage of admitted patients into four levels: no opioid use, low to medium (< 50 MME/day), high (50–90 MME/day), and very high (≥ 90 MME/day). For patients discharged home from ED, the total opioid dose administered in the ED was stratified into: no opioid use, low (< 20 MME), medium (20–50 MME), and high (≥ 50 MME). Since the total opioid dose during an ED visit was administered to patients within an average of four hours, we expected overall lower opioid doses administered in the ED compared to the mean daily dosage provided to admitted patients during hospital stay, and therefore, we separately reported low and medium level of opioid doses during an ED visit. We described the opioid dose of patients discharged home from ED and the daily opioid dosage of admitted patients by demographic characteristics and features of the ED visit (and hospital admission) and calculated the mean and standard deviation (SD). The daily opioid dosage of admitted patients was also separately analysed by the category of inpatient diagnosis. We fitted multivariate logistic regression models and used both-direction stepwise model selection method by the Akaike information criterion (AIC) value [22] to identify the best fit model and factors associated with high daily opioid dosages of ≥ 50 MME/day [3] in admitted patients who received opioid therapy during hospital stay. Potential factors included in the models were age as a continuous variable and gender, socio-economic status, country of birth, preferred language, interpreter use, inpatient diagnosis, ambulance use, triage category, and length of hospital admission as binary variables. Inpatient diagnoses were dichotomised to serious pathology and non-serious conditions. We used Hosmer-Lemeshow test and C-statistics to check the fit and apparent performance of the model [23]. Subgroup analyses were conducted to create modelling of high daily dosage by two age groups, i.e., patients aged < 65 years and ≥ 65 years, given their clinical relevance to opioid use practices and the risk of opioid-related adverse events. The break point at 65 years was verified by the Davies’ test and the break-point estimate of the segmented relationship between age and high daily dosage. Results We identified 9,345 eligible ED encounters with an ED diagnosis of non-specific low back pain or radiculopathy from 1 January 2016 to 31 October 2021, including 7,633 patients (81.7%) who were discharged home and 1,712 (18.3%) admitted to hospitals. In total, we analysed 12,850 opioid orders of 10 types of active opioid ingredient placed for 5,132 patients (54.9%) in the ED or during hospital admission. Patients discharged home from ED The opioid dose administered in the ED to patients who were discharged home is described in Table 1. Among the 7,633 patients, 3,519 patients (46.1%) received an opioid at an average dose of 13.9 MME (SD = 35.6), including 3,024 (39.6%) received < 20 MME, 426 (5.6%) received 20–50 MME, and 69 (0.9%) received ≥ 50 MME. The percentage of patients who received opioid doses of ≥ 50 MME in the ED was higher in those triaged as urgent cases (3.2%), had a prolonged stay in the ED (2.1%), and arrived by ambulance (1.7%). Table 1Individual opioid dose administered in ED patients with low back pain who were discharged homeN (%)No. of patients0 MME< 20 MME20–50 MME≥ 50 MMEMean dose (SD)aOverall7,6334,114 (53.9)3,024 (39.6)426 (5.6)69 (0.9)13.9 (35.6)Non-specific low back pain6,8013,728 (54.8)2,632 (38.7)380 (5.6)61 (0.9)14.2 (37.8)Radiculopathy832386 (46.4)392 (47.1)46 (5.5)8 (1.0)11.9 (10.7)Age Mean age (SD), years50.3 (19.0)49.3 (18.9)50.8 (19.1)50.6 (17.7)47.7 (15.3)- Adult (18–64 years)5,7133,135 (54.9)2,204 (38.6)316 (5.5)58 (1.0)14.2 (24.4) Older adult (≥ 65 years)1,920979 (51.0)820 (42.7)110 (5.7)11 (0.6)13.1 (55.7)Gender Female3,7982,005 (52.8)1,557 (41.0)197 (5.2)39 (1.0)13.6 (25.1) Male3,8342,108 (55.0)1,467 (38.3)229 (6.0)30 (0.8)14.2 (43.9) Indeterminate11 (100)0Socio-economic status Advantaged5,8913,076 (52.2)2,396 (40.7)360 (6.1)59 (1.0)14.4 (38.8) Disadvantaged1,253732 (58.4)464 (37.0)49 (3.9)8 (0.6)12.5 (19.3) Residing overseas489306 (62.6)164 (33.5)17 (3.5)2 (0.4)10.9 (8.1)Country of birth Australia3,7351,947 (52.1)1,518 (40.6)226 (6.1)44 (1.2)15.4 (47.0) Overseas3,8982,167 (55.6)1,506 (38.6)200 (5.1)25 (0.6)12.4 (16.9)Preferred language English5,9323,153 (53.2)2,371 (40.0)346 (5.8)62 (1.0)14.6 (39.7) Non-English1,701961 (56.5)653 (38.4)80 (4.7)7 (0.4)11.3 (9.0)Required an interpreter No7,1353,835 (53.7)2,829 (39.6)404 (5.7)67 (0.9)14.1 (36.7) Yes498279 (56.0)195 (39.2)22 (4.4)2 (0.4)10.5 (7.9)Mode of transport Ambulance2,052967 (47.1)863 (42.1)188 (9.2)34 (1.7)17.7 (58.0) Self-presented5,5813,147 (56.4)2,161 (38.7)238 (4.3)35 (0.6)12.2 (18.0)Triage category Not urgent (category 3–5)7,5084,059 (54.1)2,974 (39.6)410 (5.5)65 (0.9)13.8 (35.7) Urgent (category 1–2)12555 (44.0)50 (40.0)16 (12.8)4 (3.2)21.1 (29.6)ED length of stay Mean length (SD), hours3.6 (2.1)3.1 (1.8)3.9 (2.2)5.0 (2.9)5.3 (3.1)- Shorter stay (≤ 4 h)6,2083,574 (57.6)2,329 (37.5)266 (4.3)39 (0.6)12.2 (17.4) Prolonged stay (> 4 h)1,425540 (37.9)695 (48.8)160 (11.2)30 (2.1)18.9 (64.0)Time of ED arrival Working hours3,0541,752 (57.4)1,136 (37.2)146 (4.8)20 (0.7)13.7 (49.3) After hours4,5792,362 (51.6)1,888 (41.2)280 (6.1)49 (1.1)14.1 (24.1)a Mean opioid dose (SD) of patients who were administered an opioid during an ED visit Table 1Individual opioid dose administered in ED patients with low back pain who were discharged homeN (%)No. of patients0 MME< 20 MME20–50 MME≥ 50 MMEMean dose (SD)aOverall7,6334,114 (53.9)3,024 (39.6)426 (5.6)69 (0.9)13.9 (35.6)Non-specific low back pain6,8013,728 (54.8)2,632 (38.7)380 (5.6)61 (0.9)14.2 (37.8)Radiculopathy832386 (46.4)392 (47.1)46 (5.5)8 (1.0)11.9 (10.7)Age Mean age (SD), years50.3 (19.0)49.3 (18.9)50.8 (19.1)50.6 (17.7)47.7 (15.3)- Adult (18–64 years)5,7133,135 (54.9)2,204 (38.6)316 (5.5)58 (1.0)14.2 (24.4) Older adult (≥ 65 years)1,920979 (51.0)820 (42.7)110 (5.7)11 (0.6)13.1 (55.7)Gender Female3,7982,005 (52.8)1,557 (41.0)197 (5.2)39 (1.0)13.6 (25.1) Male3,8342,108 (55.0)1,467 (38.3)229 (6.0)30 (0.8)14.2 (43.9) Indeterminate11 (100)0Socio-economic status Advantaged5,8913,076 (52.2)2,396 (40.7)360 (6.1)59 (1.0)14.4 (38.8) Disadvantaged1,253732 (58.4)464 (37.0)49 (3.9)8 (0.6)12.5 (19.3) Residing overseas489306 (62.6)164 (33.5)17 (3.5)2 (0.4)10.9 (8.1)Country of birth Australia3,7351,947 (52.1)1,518 (40.6)226 (6.1)44 (1.2)15.4 (47.0) Overseas3,8982,167 (55.6)1,506 (38.6)200 (5.1)25 (0.6)12.4 (16.9)Preferred language English5,9323,153 (53.2)2,371 (40.0)346 (5.8)62 (1.0)14.6 (39.7) Non-English1,701961 (56.5)653 (38.4)80 (4.7)7 (0.4)11.3 (9.0)Required an interpreter No7,1353,835 (53.7)2,829 (39.6)404 (5.7)67 (0.9)14.1 (36.7) Yes498279 (56.0)195 (39.2)22 (4.4)2 (0.4)10.5 (7.9)Mode of transport Ambulance2,052967 (47.1)863 (42.1)188 (9.2)34 (1.7)17.7 (58.0) Self-presented5,5813,147 (56.4)2,161 (38.7)238 (4.3)35 (0.6)12.2 (18.0)Triage category Not urgent (category 3–5)7,5084,059 (54.1)2,974 (39.6)410 (5.5)65 (0.9)13.8 (35.7) Urgent (category 1–2)12555 (44.0)50 (40.0)16 (12.8)4 (3.2)21.1 (29.6)ED length of stay Mean length (SD), hours3.6 (2.1)3.1 (1.8)3.9 (2.2)5.0 (2.9)5.3 (3.1)- Shorter stay (≤ 4 h)6,2083,574 (57.6)2,329 (37.5)266 (4.3)39 (0.6)12.2 (17.4) Prolonged stay (> 4 h)1,425540 (37.9)695 (48.8)160 (11.2)30 (2.1)18.9 (64.0)Time of ED arrival Working hours3,0541,752 (57.4)1,136 (37.2)146 (4.8)20 (0.7)13.7 (49.3) After hours4,5792,362 (51.6)1,888 (41.2)280 (6.1)49 (1.1)14.1 (24.1)a Mean opioid dose (SD) of patients who were administered an opioid during an ED visit Individual opioid dose administered in ED patients with low back pain who were discharged home a Mean opioid dose (SD) of patients who were administered an opioid during an ED visit Patients admitted to hospitals The daily opioid dosage for admitted patients, including both opioids administered in the ED and during hospital admission, are presented in Table 2. Almost 95% of patients (1,613 out of 1,712) admitted to hospital with a provisional ED diagnosis of non-serious low back pain were provided opioid therapy during hospital stay at a mean dosage of 34.7 MME/day (SD = 68.3), including 154 patients (9.0%) receiving 50–90 MME/day and 117 (6.8%) receiving ≥ 90 MME/day. Table 2Characteristics and the daily opioid dosage of patients who were admitted to hospitalsN (%)All admitted patientsAdmitted patients with non-serious low back painbNo. of patients0 MME/day< 50 MME/day50–90 MME/day≥ 90 MME/dayMean dosage (SD)aNo. of patients0 MME/day< 50 MME/day50–90 MME/day≥ 90 MME/dayMean dosage (SD)aOverall1,71299 (5.8)1,342 (78.4)154 (9.0)117 (6.8)34.7 (68.3)96842 (4.3)788 (81.4)92 (9.5)46 (4.8)32.3 (67.1)Age Mean age (SD), years67.8 (18.7)64.7 (20.3)69.6 (18.6)62.3 (18.4)60.7 (14.9)-65.1 (19.1)62.5 (19.0)66.5 (19.1)59.5 (18.8)56.5 (15.6)- Adult (18–64 years)62947 (7.5)435 (69.2)78 (12.4)69 (11.0)49.2 (92.7)42024 (5.7)313 (74.5)55 (13.1)28 (6.7)40.6 (77.0) Older adult (≥ 65 years)1,08352 (4.8)907 (83.7)76 (7.0)48 (4.4)26.5 (47.7)54818 (3.3)475 (86.7)37 (6.8)18 (3.3)26.1 (57.9)Gender Female1,02163 (6.2)821 (80.4)86 (8.4)51 (5.0)29.3 (54.5)57128 (4.9)474 (83.0)50 (8.8)19 (3.3)28.1 (58.7) Male69136 (5.2)521 (75.4)68 (9.8)66 (9.6)42.7 (84.0)39714 (3.5)314 (79.1)42 (10.6)27 (6.8)38.3 (77.3)Socio-economic status Advantaged1,43491 (6.3)1,136 (79.2)120 (8.4)87 (6.1)32.8 (33.0)79938 (4.8)663 (83.0)70 (8.8)28 (3.5)29.7 (65.3) Disadvantaged2606 (2.3)193 (74.2)33 (12.7)28 (10.8)40.2 (55.2)1574 (2.5)116 (73.9)21 (13.4)16 (10.2)38.8 (47.8) Residing overseas182 (11.1)13 (72.2)1 (5.6)2 (11.1)91.3 (184.4)120 (0)9 (75.0)1 (8.3)2 (16.7)112.6 (210.4)Country of birth Australia76560 (7.8)574 (75.0)69 (9.0)62 (8.1)41.4 (92.0)42725 (5.9)337 (78.9)39 (9.1)26 (6.1)40.3 (95.2) Overseas94739 (4.1)768 (81.1)85 (9.0)55 (5.8)29.5 (40.8)54117 (3.1)451 (83.4)53 (9.8)20 (3.7)26.2 (30.6)Preferred language English1,14276 (6.7)874 (76.5)103 (9.0)89 (7.8)38.5 (80.3)65536 (5.5)523 (79.8)60 (9.2)36 (5.5)36.1 (79.8) Non-English57023 (4.0)468 (82.1)51 (8.9)28 (4.9)27.3 (33.4)3136 (1.9)265 (84.7)32 (10.2)10 (3.2)24.6 (25.8)Required an interpreter No1,46487 (5.9)1,133 (77.4)135 (9.2)109 (7.4)36.5 (72.7)82840 (4.8)663 (80.1)81 (9.8)44 (5.3)34.3 (72.0) Yes24812 (4.8)209 (84.3)19 (7.7)8 (3.2)24.2 (31.6)1402 (1.4)125 (89.3)11 (7.9)2 (1.4)21.1 (22.1)Mode of transport Ambulance1,01652 (5.1)815 (80.2)85 (8.4)64 (6.3)33.2 (72.4)58125 (4.3)481 (82.8)50 (8.6)25 (4.3)30.1 (70.7) Self-presented69647 (6.8)527 (75.7)69 (9.9)53 (7.6)36.9 (61.8)38717 (4.4)307 (79.2)42 (10.9)21 (5.4)35.7 (61.3)Triage category Not urgent (category 3–5)1,64992 (5.6)1,299 (78.0)146 (8.9)112 (6.8)34.8 (69.2)93741 (4.4)761 (81.2)90 (9.6)45 (4.8)32.7 (68.1) Urgent (category 1–2)637 (11.1)43 (68.3)8 (12.7)5 (7.9)33.4 (36.7)311 (3.2)27 (87.1)2 (6.5)1 (3.2)21.8 (22.4)Mean length of hospital stay Mean length (SD), days9.6 (12.7)6.0 (11.1)10.0 (11.4)9.4 (9.7)13.8 (25.5)-6.8 (7.4)4.5 (9.9)7.3 (7.5)7.5 (7.5)5.1 (4.3)- Shorter admission (≤ 6 days)90780 (8.8)688 (75.9)78 (8.6)61 (6.7)33.7 (62.1)60939 (6.4)481 (79.0)54 (8.9)35 (5.7)33.8 (67.0) Prolonged admission (> 6 days)80519 (2.4)654 (81.2)76 (9.4)56 (7.0)35.8 (74.4)3593 (0.8)307 (85.5)38 (10.6)11 (3.1)29.9 (67.4)a Mean daily opioid dosage (SD) in patients who received opioid therapy during hospital stayb The principal diagnosis at inpatient discharge Table 2Characteristics and the daily opioid dosage of patients who were admitted to hospitalsN (%)All admitted patientsAdmitted patients with non-serious low back painbNo. of patients0 MME/day< 50 MME/day50–90 MME/day≥ 90 MME/dayMean dosage (SD)aNo. of patients0 MME/day< 50 MME/day50–90 MME/day≥ 90 MME/dayMean dosage (SD)aOverall1,71299 (5.8)1,342 (78.4)154 (9.0)117 (6.8)34.7 (68.3)96842 (4.3)788 (81.4)92 (9.5)46 (4.8)32.3 (67.1)Age Mean age (SD), years67.8 (18.7)64.7 (20.3)69.6 (18.6)62.3 (18.4)60.7 (14.9)-65.1 (19.1)62.5 (19.0)66.5 (19.1)59.5 (18.8)56.5 (15.6)- Adult (18–64 years)62947 (7.5)435 (69.2)78 (12.4)69 (11.0)49.2 (92.7)42024 (5.7)313 (74.5)55 (13.1)28 (6.7)40.6 (77.0) Older adult (≥ 65 years)1,08352 (4.8)907 (83.7)76 (7.0)48 (4.4)26.5 (47.7)54818 (3.3)475 (86.7)37 (6.8)18 (3.3)26.1 (57.9)Gender Female1,02163 (6.2)821 (80.4)86 (8.4)51 (5.0)29.3 (54.5)57128 (4.9)474 (83.0)50 (8.8)19 (3.3)28.1 (58.7) Male69136 (5.2)521 (75.4)68 (9.8)66 (9.6)42.7 (84.0)39714 (3.5)314 (79.1)42 (10.6)27 (6.8)38.3 (77.3)Socio-economic status Advantaged1,43491 (6.3)1,136 (79.2)120 (8.4)87 (6.1)32.8 (33.0)79938 (4.8)663 (83.0)70 (8.8)28 (3.5)29.7 (65.3) Disadvantaged2606 (2.3)193 (74.2)33 (12.7)28 (10.8)40.2 (55.2)1574 (2.5)116 (73.9)21 (13.4)16 (10.2)38.8 (47.8) Residing overseas182 (11.1)13 (72.2)1 (5.6)2 (11.1)91.3 (184.4)120 (0)9 (75.0)1 (8.3)2 (16.7)112.6 (210.4)Country of birth Australia76560 (7.8)574 (75.0)69 (9.0)62 (8.1)41.4 (92.0)42725 (5.9)337 (78.9)39 (9.1)26 (6.1)40.3 (95.2) Overseas94739 (4.1)768 (81.1)85 (9.0)55 (5.8)29.5 (40.8)54117 (3.1)451 (83.4)53 (9.8)20 (3.7)26.2 (30.6)Preferred language English1,14276 (6.7)874 (76.5)103 (9.0)89 (7.8)38.5 (80.3)65536 (5.5)523 (79.8)60 (9.2)36 (5.5)36.1 (79.8) Non-English57023 (4.0)468 (82.1)51 (8.9)28 (4.9)27.3 (33.4)3136 (1.9)265 (84.7)32 (10.2)10 (3.2)24.6 (25.8)Required an interpreter No1,46487 (5.9)1,133 (77.4)135 (9.2)109 (7.4)36.5 (72.7)82840 (4.8)663 (80.1)81 (9.8)44 (5.3)34.3 (72.0) Yes24812 (4.8)209 (84.3)19 (7.7)8 (3.2)24.2 (31.6)1402 (1.4)125 (89.3)11 (7.9)2 (1.4)21.1 (22.1)Mode of transport Ambulance1,01652 (5.1)815 (80.2)85 (8.4)64 (6.3)33.2 (72.4)58125 (4.3)481 (82.8)50 (8.6)25 (4.3)30.1 (70.7) Self-presented69647 (6.8)527 (75.7)69 (9.9)53 (7.6)36.9 (61.8)38717 (4.4)307 (79.2)42 (10.9)21 (5.4)35.7 (61.3)Triage category Not urgent (category 3–5)1,64992 (5.6)1,299 (78.0)146 (8.9)112 (6.8)34.8 (69.2)93741 (4.4)761 (81.2)90 (9.6)45 (4.8)32.7 (68.1) Urgent (category 1–2)637 (11.1)43 (68.3)8 (12.7)5 (7.9)33.4 (36.7)311 (3.2)27 (87.1)2 (6.5)1 (3.2)21.8 (22.4)Mean length of hospital stay Mean length (SD), days9.6 (12.7)6.0 (11.1)10.0 (11.4)9.4 (9.7)13.8 (25.5)-6.8 (7.4)4.5 (9.9)7.3 (7.5)7.5 (7.5)5.1 (4.3)- Shorter admission (≤ 6 days)90780 (8.8)688 (75.9)78 (8.6)61 (6.7)33.7 (62.1)60939 (6.4)481 (79.0)54 (8.9)35 (5.7)33.8 (67.0) Prolonged admission (> 6 days)80519 (2.4)654 (81.2)76 (9.4)56 (7.0)35.8 (74.4)3593 (0.8)307 (85.5)38 (10.6)11 (3.1)29.9 (67.4)a Mean daily opioid dosage (SD) in patients who received opioid therapy during hospital stayb The principal diagnosis at inpatient discharge Characteristics and the daily opioid dosage of patients who were admitted to hospitals a Mean daily opioid dosage (SD) in patients who received opioid therapy during hospital stay b The principal diagnosis at inpatient discharge Daily opioid dosage varied significantly by age group, gender, and socio-economic status. Admitted patients aged < 65 years received an average dosage of 49.2 MME/day (SD = 92.7), including 23.4% (147 out of 629 patients) received ≥ 50 MME/day; whereas the mean dosage in those ≥ 65 years was 26.5 MME/day (SD = 47.7). Male patients (42.7 MME/day, SD = 84.0) on average received a higher daily dosage than females (29.3 MME/day, SD = 54.5). Almost 25% of patients (61 out of 260) residing in disadvantaged suburbs received ≥ 50 MME/day, compared to 14.5% (207 out of 1,434) of those residing in advantaged suburbs. Admitted patients by inpatient diagnosis The daily opioid dosages of admitted patients categorised by the principal diagnosis at inpatient discharge is described in Table 3. On average, admitted patients who were subsequently found to have serious pathology during hospital admission received a higher daily opioid dosage and stayed in hospital for twice as long (40.5 MME/day; 14.8 days) compared to patients with a non-serious condition (31.4 MME/day; 7.1 days). Among the 968 admitted patients with an inpatient diagnosis of non-specific low back pain or radiculopathy (Table 2), 926 patients (95.7%) received opioids during hospital stay, including 92 patients (9.5%) received 50–90 MME/day and 46 (4.8%) received ≥ 90 MME/day. Table 3The daily opioid dosage of patients admitted to hospitals (n = 1,712) categorised by the principal diagnosis at inpatient dischargeNo. of patients (%)Mean length of stay (days)0 MME/day,N (%)< 50 MME/day, N (%)50–90 MME/day, N (%)≥ 90 MME/day, N (%)Mean dosage (SD)aNon-specific low back pain440 (25.7)6.425 (5.7)368 (83.6)29 (6.6)18 (4.1)31.9 (78.1)Radiculopathy528 (30.8)7.717 (3.2)420 (79.5)63 (11.9)28 (5.3)32.7 (56.8)Spinal serious pathologyb224 (13.1)13.712 (5.4)171 (76.3)25 (11.2)16 (7.1)35.5 (58.2)Serious pathology beyond lumbar spinec418 (24.4)15.438 (9.1)300 (71.8)29 (6.9)51 (12.1)43.2 (81.6)Non-lumbar musculoskeletal conditionsd102 (6.0)7.37 (6.9)82 (81.4)8 (7.8)4 (3.9)22.5 (27.7)a Mean daily opioid dosage (SD) in patients who received opioid therapy during hospital stayb Spinal serious pathology: e.g. fracture of thoracic/lumbar vertebra, lumbar osteomyelitis, lumbosacral region discitisc Serious pathology beyond lumbar spine: e.g. malignant melanoma of trunk, cellulitis of lower limb, acute kidney failured Non-lumbar musculoskeletal conditions: e.g. musculoskeletal pain in hip, primary coxarthrosis, trochanteric bursitis Table 3The daily opioid dosage of patients admitted to hospitals (n = 1,712) categorised by the principal diagnosis at inpatient dischargeNo. of patients (%)Mean length of stay (days)0 MME/day,N (%)< 50 MME/day, N (%)50–90 MME/day, N (%)≥ 90 MME/day, N (%)Mean dosage (SD)aNon-specific low back pain440 (25.7)6.425 (5.7)368 (83.6)29 (6.6)18 (4.1)31.9 (78.1)Radiculopathy528 (30.8)7.717 (3.2)420 (79.5)63 (11.9)28 (5.3)32.7 (56.8)Spinal serious pathologyb224 (13.1)13.712 (5.4)171 (76.3)25 (11.2)16 (7.1)35.5 (58.2)Serious pathology beyond lumbar spinec418 (24.4)15.438 (9.1)300 (71.8)29 (6.9)51 (12.1)43.2 (81.6)Non-lumbar musculoskeletal conditionsd102 (6.0)7.37 (6.9)82 (81.4)8 (7.8)4 (3.9)22.5 (27.7)a Mean daily opioid dosage (SD) in patients who received opioid therapy during hospital stayb Spinal serious pathology: e.g. fracture of thoracic/lumbar vertebra, lumbar osteomyelitis, lumbosacral region discitisc Serious pathology beyond lumbar spine: e.g. malignant melanoma of trunk, cellulitis of lower limb, acute kidney failured Non-lumbar musculoskeletal conditions: e.g. musculoskeletal pain in hip, primary coxarthrosis, trochanteric bursitis The daily opioid dosage of patients admitted to hospitals (n = 1,712) categorised by the principal diagnosis at inpatient discharge a Mean daily opioid dosage (SD) in patients who received opioid therapy during hospital stay b Spinal serious pathology: e.g. fracture of thoracic/lumbar vertebra, lumbar osteomyelitis, lumbosacral region discitis c Serious pathology beyond lumbar spine: e.g. malignant melanoma of trunk, cellulitis of lower limb, acute kidney failure d Non-lumbar musculoskeletal conditions: e.g. musculoskeletal pain in hip, primary coxarthrosis, trochanteric bursitis Modelling of high opioid dosage in admitted patients Among admitted patients who received opioid therapy during hospital stay, age, gender, socio-economic status, and inpatient diagnosis were associated with high daily opioid dosages of ≥ 50 MME/day. The adjusted odds of receiving ≥ 50 MME/day in the hospital were 0.80 (95% CI 0.74, 0.85) times lower per 10 years increase in age, 1.38 (95% CI 1.05, 1.81) times higher in males than females, 1.50 (95% CI 1.07, 2.10) times higher in patients residing in disadvantaged suburbs than those from advantaged suburbs, and 1.79 (95% CI 1.36, 2.37) times higher in patients diagnosed with serious pathology during hospital admission than those with a non-serious condition. Other potential factors were not included in the final model. Hosmer-Lemeshow test (χ2 = 7.85, df = 8, p = 0.45) showed no lack of model fit, and the C-statistic of 0.67 (95%CI 0.64, 0.71) indicated reasonable discrimination. Subgroup analysis in admitted patients aged ≥ 65 years (n = 1,083) showed that the adjusted odds of receiving an opioid dosage of ≥ 50 MME/day was 0.53 (95% CI 0.41, 0.68) times lower per 10 years increase in age and 1.61 (95% CI 1.09, 2.36) times higher in patients with an inpatient diagnosis of serious pathology compared with those with a non-serious condition; whereas gender and socio-economic status were adjusted for but showed no significant effects. In patients aged < 65 years (n = 629), after adjusting for length of hospital admission, the odds of receiving high daily opioid dosage were 2.08 (95% CI 1.37, 3.16) times higher in patients with serious pathology. Discussion This study examined the opioid dose for low back pain during an ED visit and the average daily opioid dosage in patients admitted to hospitals. The opioid dose administered to ED patients discharged home was higher in more urgent cases, patients who arrived by ambulance, and those who stayed longer in the ED. Almost 95% of admitted patients with low back pain received opioid therapy, with 15.8% of those receiving a high dosage of ≥ 50 MME/day. High-dose opioid therapy was associated with inpatient diagnosis of serious pathology, younger age, male gender, and disadvantaged socio-economic status. Subgroup analysis showed that age only modified opioid dosage in older patients aged ≥ 65 years. Gender differences in opioid dosage have been reported in previous studies. In primary care, male patients had a higher risk of being prescribed opioids of ≥ 90 MME/day than females (relative risk = 1.21, 95% CI 1.14, 1.28) [24]. Another review reported no pooled gender difference in the response to opioids for acute pain (mean difference = 0.42 MME, 95% CI -0.07, 0.91), but males with chronic non-cancer pain were prescribed higher daily opioid dosages in the > 45 years age group (mean difference = 24.89 MME/day, 95% CI 7.91, 41.87) [25]. Similarly, our study also found that male patients with low back pain were more likely than females to receive high daily opioid dosage during hospital stay, but the percentage of receiving any opioid therapy showed no significant gender difference. As opioids have weight-dependent body distribution and clearance, the gender difference could be attributed to body weight difference, which was not adjusted for in our analysis. Pain management in older patients is challenging due to increased risks with both non-opioid pharmacologic therapies and opioid therapy. Older patients have increased vulnerability to accumulation of medications, potential drug-drug interactions, and a narrower therapeutic window of safe dosage [3]. They are also at greater risk of more serious adverse events and long-term opioid use after opioid initiation [8, 26]. Our study found similar opioid doses administered to younger and older ED patients discharged home, whereas the mean daily dosage of older patients admitted to hospitals was only half of the dosage in the younger group. Despite receiving, on average, a medium level of daily opioid dosage during hospital stay, evidence showed limited benefits of opioid therapy for moderate to severe chronic low back pain in older patients [8]. Therefore, hospital clinicians should carefully consider the risks and benefits before initiating opioid therapy for older patients in the hospital, have a deprescribing plan developed and implemented once initiated [27], and encourage follow-up reviews after discharge to monitor effectiveness and safety if opioid therapy is prolonged. Our results suggest the need for more cautious use of opioids for patients with low back pain in Australian hospitals. Although we found relatively low-dose opioid therapy administered to half of ED patients with non-serious low back pain discharged home, there is clear evidence suggesting that non-opioid analgesics, such as NSAIDs, may provide similar benefits while avoiding the risk of long-term opioid use in the community [3, 12]. Clinical guidelines [4] recommend that hospital clinicians should restrict opioid therapy to admitted patients with severe pain that has not responded to non-opioid therapy, whereas mild to moderate pain should be managed with non-opioid analgesics. However, over 95% of admitted patients with an inpatient diagnosis of non-specific low back pain or radiculopathy in our study received opioid therapy during hospital stay. Evidence-based opioid sparing strategies are required in the hospital setting to reduce high-dose opioid therapy for low back pain. Our findings indicated that more than 15% of admitted patients received opioid therapy exceeding a daily dosage of 50 MME/day during hospital stay, including 5% who received very high dosages of ≥ 90 MME/day. To reduce opioid dosage in the hospital, opioid therapy for severe pain should be initiated with a weak opioid in combination with simple analgesic, such as codeine plus paracetamol or ibuprofen, before progressing to stronger and single-ingredient opioids such as immediate-release oxycodone. The synergistic effects of the combinations of opioids and simple analgesics on osteoarthritis pain [28] may also apply to low back pain, which would reduce the need for an increased dosage of single ingredient strong opioids. Using non-opioid analgesics and non-pharmaceutical therapies, such as NSAIDs and heat wrap, have been shown as effective opioid sparing strategies for low back pain in the ED when combined with clinician education and audit and feedback; [29] whereas no previous research reported effective strategies in the inpatient setting. Vulnerable populations, such as older patients, may need special considerations when developing these strategies. We were unable to access the medication administration records to identify the actual opioid doses administered to patients in the hospital and count the number of days of opioid use during hospital stay. Therefore, we were at risk of underestimating the daily opioid dosage by averaging the dosage across days without any opioid use. The completeness of medical records and inconsistency between data sources may also have impacted our results. For example, over 1,100 ED encounters had conflicting records of opioid use between data sources and were therefore excluded from analysis, which might have underestimated the overall percentage of opioid use by up to 10%. Besides, we were unable to obtain and adjust for data on patient-reported outcomes and medical history, such as pain severity and history of opioid use, which were associated with the initiation and dosage of opioid therapy in the hospital. Opioid therapy is almost universally administered to admitted patients with low back pain in Australian hospitals, and 15.8% of these patients receive a high dosage of ≥ 50 MME/day. High opioid dosage is associated with inpatient diagnosis of serious pathology, younger age < 65 years, male gender, and disadvantaged socio-economic status. Future studies investigating opioid sparing strategies to reduce the use of opioids for low back pain in the hospital setting, especially minimising high-dose opioid therapy for admitted patients, are warranted. Supplementary Information Supplementary Material 1. Appendix tables. Supplementary Material 1. Appendix tables.