Search

Background: Knowledge gaps exist around diagnostic and treatment approaches for patients admitted to hospital with low back pain. Methods: Medical record review of patients admitted to three Sydney teaching hospitals with a provisional emergency department diagnosis of non‐serious low back pain, from 2016 to 2020. Data on demographic variables, hospital costs, length of stay (LOS), diagnostic imaging and analgesic administration were extracted. Logistic regression was used to identify predictors of longer hospital stay, advanced imaging, and concomitant use of sedating medicines. Results: Median inpatient LOS for non‐specific low back pain was 4 days (interquartile range [IQR] 2‐7), and for radicular low back pain was 4 days (IQR 3‐10). Older patients with non‐serious low back pain were more likely to stay longer, as were arrivals by ambulance. Plain lumbar radiography was used in 8.3% of admissions, whereas 37.6% of patients received advanced lumbar imaging (computed tomography or magnetic resonance imaging). Opioids were administered in ~80% of admissions; 49% of patients with radicular low back pain were given an antiepileptic in addition to an opioid. In all, 18.4% of admissions resulted in at least one hospital‐acquired complication, such as an accidental fall (3.1%) or a medication‐related adverse effect (13.3%). Physiotherapists saw 82.6% of low back pain admissions. Costs of inpatient care were estimated at a mean of AU$ 14 000 per admission. Conclusions: We noted relatively high rates of concomitant use of sedating pain medicines and referrals for advanced lumbar imaging and laboratory tests. Strategies to address these issues in inpatient care of low back pain are needed. [ABSTRACT FROM AUTHOR]

To determine the proportion of patients admitted to the hospital for back pain who have nonserious back pain, serious spinal, or serious other pathology as their final diagnosis. The proportion of nonserious back pain admissions will be used to plan for future 'virtual hospital' admissions. Electronic medical record data between January 2016 and September 2020 from three emergency departments (ED) in Sydney, Australia were used to identify inpatient admissions. SNOMED-CT-AU diagnostic codes were used to select ED patients aged 18 and older with an admitting diagnosis related to nonserious back pain. The inpatient discharge diagnosis was determined from the primary ICD-10-AM codes by two independent clinician-researchers. Inpatient admissions were then analysed by sociodemographic and hospital admission variables. A total of 38.1% of patients admitted with a provisional diagnosis of nonserious back pain were subsequently diagnosed with a specific pathology likely unsuitable for virtual care; 14.2% with a serious spinal pathology (e.g., fracture and infection) and 23.9% a serious pathology beyond the lumbar spine (e.g., pathological fracture and neoplasm). A total of 57% of admissions were identified as nonserious back pain, likely suitable for virtual care. A challenge for implementing virtual care in this setting is screening for patients with serious pathology. Protocols need to be developed to reduce the risk of patients being admitted to virtual hospitals with serious pathology as the cause of their back pain. Key Points • Among admitted patients provisionally diagnosed in ED with non-serious back pain, 38.1% were found to have 'serious spinal pathologies' or 'serious pathologies beyond the lumbar spine' at discharge. • Spinal fractures were the most common serious spinal pathology, accounting for 9% of all provisional 'non-serious back pain' admissions from ED. • 57% of back pain admissions were confirmed to be non-serious back pain and may be suitable to virtual hospital care; the challenge is discriminating these patients from those with serious pathology. [ABSTRACT FROM AUTHOR]