Low-Dose Morphine Intrathecal Analgesia in Elderly Patients with Hip Fracture Undergoing Single Spinal Anesthesia: A Randomized Controlled Trial

Introduction

Hip fracture (HF) predominantly affects the elderly population, with over 40% of cases occurring in individuals over 85 years of age.¹ In 2019, HFs affected 14.2 million people worldwide.² As the population ages, HF imposes a growing socioeconomic burden, and its incidence is expected to nearly double by 2050.¹ After surgical HF repair, effective postoperative analgesia is essential for early rehabilitation and mobilization; however, pain is often undertreated, yielding worse outcomes, prolonged hospital stays, and higher healthcare costs.³ Moreover, among elderly patients, insufficiently managed postoperative pain is associated with increased risk of postoperative delirium.^4,5 Despite these consequences, there are currently no evidence-based guidelines for postoperative pain management after surgical HF repair. Standard care includes multimodal analgesia, with emphasis on minimizing systemic opioid use and avoiding NSAIDs.^6,7 In contrast, there are well-established perioperative care guidelines for total hip arthroplasty (THA).^8,9

Analgesic strategies for THA have been extensively evaluated by the Procedure-Specific Postoperative Pain Management (PROSPECT) group. In the elective setting, intrathecal morphine is recommended as part of multimodal analgesia.¹⁰ Although THA is often the surgical procedure of choice for hip fracture, randomized evidence specifically addressing the use of ITM in the acute hip fracture population remains limited. Given the aging population and the increasing incidence of hip fractures, optimizing safe and effective analgesia in this setting constitutes a clinically relevant priority.

Multiple analgesic strategies are available for patients after HF surgery, including systemic analgesia, patient-controlled intravenous analgesia, multimodal pain management, and regional techniques, such as intrathecal opioid administration and peripheral nerve blocks (PNBs). PNBs, such as the fascia iliaca compartment block and femoral nerve block, represent established regional techniques with a favourable safety profile.¹⁰ However, their use is associated with additional procedural time and requires specific expertise in ultrasound-guided regional anesthesia. In contrast, in patients undergoing HF repair under single spinal anesthesia, intrathecal morphine (ITM) offers an analgesic modality characterized by rapid and relatively simple administration, a low failure rate, cost-effectiveness, and minimal equipment or postoperative care requirements.

ITM is a well-established method of perioperative analgesia which provides pain relief for 20–48 hours.^11,12 Current evidence indicates that low-dose ITM (75–150 µg) is associated with limited adverse effects.^3,13 In patients undergoing elective THA, previous studies have identified 100 μg as an optimal balance between effective analgesia and the risk of adverse effects when compared with both lower (50 μg) and higher (200 μg) doses.^14–16 Consistent with these dose-response data, 100 μg has been shown to provide effective analgesia while limiting the risk of dose-dependent adverse effects.

Anger et al⁹ suggest that ITM should be considered when THA is performed with spinal anesthesia. However, only limited evidence is available regarding ITM in elderly patients undergoing acute hip osteosynthesis, and concerns have been raised about this population’s increased susceptibility to adverse effects, particularly opioid-related respiratory depression. Age over 65 years is considered an independent risk factor for respiratory depression.¹⁷ Other patient-related risk factors for respiratory depression include acute or chronic conditions that affect the level of consciousness, hepatic or renal function (metabolism), plasma protein binding (eg, hypoalbuminemia), chemosensitivity of the respiratory center, and individual opioid tolerance.¹⁸ Importantly, most adverse effects can be reduced by using a low-dose regimen, and pruritus and postoperative nausea and vomiting (PONV) can be further reduced through early administration of a 5-HT3 antagonist (ondansetron).¹⁹

This parallel-group superiority trial was performed with the aim of comparing the efficacy and safety of 100 µg ITM versus standard multimodal analgesia among elderly patients undergoing acute HF surgery under single spinal anesthesia. The study design incorporates contemporary strategies aimed at minimizing opioid-related adverse effects. In addition to a low-dose ITM regimen, it includes the preemptive administration of 5-HT3 receptor antagonists to prevent opioid-induced pruritus and postoperative nausea and vomiting, complications which may delay ambulation, impair oral intake, and reduce patient satisfaction.⁹

The primary objective was to evaluate postoperative pain intensity, both at rest and during repositioning, within 24 hours after surgery. The secondary objectives were to assess the effects of ITM on postoperative analgesic consumption; time to first analgesic administration; and the incidence and severity of adverse effects, including respiratory depression, bradycardia, hypotension, pruritus, and PONV. Additionally, excessive sedation was assessed during routine ICU surveillance, but was not considered an outcome measure of this study. Urinary retention was not assessed because urinary catheterization was performed as standard practice during the perioperative period.

Methods

Setting

This study was approved by the Czech principal authority – State Institute for Drug Control (SÚKL; reference number: sukls266356/2022) and the Ethics Committee of University Hospital Ostrava (17. listopadu 1790/5,70852Ostrava, Czech Republic; reference number: 468/2023). The trial was registered in the EudraCT database (No. 2021–002765-17) and at ClinicalTrials.gov (NCT05920642) on June 16, 2023. Written informed consent was obtained from all participants, and the trial was reported following the CONSORT reporting guidelines.²⁰ Patients and the public had no involvement in the design or implementation of this research.

Trial Design

Each year, University Hospital Ostrava admits approximately 200 patients with HF for surgical treatment. The study investigators regularly reviewed the hospital information system to identify patients with HF, and screened these patients for study eligibility. Potential participants were verbally informed about the study by the investigators, and were provided with a written informed consent form to sign. Eligible patients were randomized into two groups. Enrollment began at the time of subarachnoid puncture in the operating room, and follow-up ended 24 hours after postoperative transfer to the intensive care unit (ICU). Progress through the phases of the trial is illustrated in the CONSORT diagram (Figure 1) and is also provided as Supplementary Figure 1.

Figure 1 CONSORT 2025 flow diagram of patient recruitment.17 Abbreviations: ITM, intrathecal morphine; NRS, numeric rating scale.

Allocation

A total of 465 patients with HF were admitted to the hospital between September 13, 2023, and July 18, 2025, reflecting a higher-than-average case volume during this interval. HF was defined as any fracture of the femur at <5 cm below the lesser trochanter. Inclusion criteria were age of 60–90 years, American Society of Anesthesiologists (ASA) classification of I–III, scheduled for single spinal anesthesia (SA), and ability to provide informed consent. Exclusion criteria were patient’s refusal to participate, contraindications to SA (eg., drug-induced and patient-related coagulopathy, injection site infection, or severe aortic stenosis), high risk of severe respiratory depression (eg., class III obesity, stage III–IV COPD, craniocerebral trauma, upper airway obstruction, etc)., and known opioid allergy. Withdrawal criteria included conversion to general anesthesia due to failed SA, use of epidural analgesia, and PNBs. Of the 465 patients with HF, 383 were not eligible because they did not meet the inclusion criteria (n = 254), declined to participate (n = 34), or met the exclusion criteria (n = 95). Thus, 82 patients were ultimately included and randomized to the intervention group (n = 47) or the control group (n = 35). Unequal group sizes resulted from variability in the number of patients enrolled per allocation week. One patient was withdrawn due to failed SA.

Randomization

Computer-generated permuted block randomization was used to pre-assign each calendar week to either the intervention or control group, with an allocation ratio of 1:1. All patients enrolled during a given week were assigned to the same group. Group allocation was determined in the operating room at the initiation of anesthetic care. Since HF requires timely surgical intervention, patients were promptly scheduled for surgery. Weekly allocation was considered the most suitable for our institutional setting, primarily due to regulatory requirements. As a controlled substance, intrathecal morphine must be stored in secured conditions and, as a sterile preservative-free solution, requires temperature control. The weekly allocation scheme facilitated safe handling and availability while avoiding delays in surgical care, particularly during out-of-hours service. In the event of imbalance in group sizes or variance, Welch’s t-test was planned as a correction. Participants were not informed of their allocation. However, double blinding was not implemented due to safety concerns, as the elderly population is particularly susceptible to opioid-related adverse events. To minimize performance and assessment bias, multiple predefined outcome measures were used, and most secondary outcomes were based on objective parameters. In addition, standardized perioperative care protocols were applied across both groups, including postoperative follow-up, pain management, and management of adverse effects.

Intervention

Upon arrival in the operating room, patient monitoring included electrocardiogram (ECG), noninvasive blood pressure (NIBP), heart rate (HR), respiratory rate (RR), and blood oxygen saturation (SpO₂). During positioning for spinal puncture, patients could receive analgesia with a small dose of ketamine IV or a short-acting opioid (alfentanil or remifentanil) IV.⁷ The lowest effective dose was used solely as necessary to facilitate patient positioning. In both groups, SA was administered with a 26-G atraumatic needle. Patients in the intervention arm received 100 µg ITM, in addition to 0.5% bupivacaine or 0.5% levobupivacaine. Levobupivacaine was predominantly used due to its favorable safety profile in elderly patients and institutional preference. The choice was at the discretion of the attending anesthesiologist and was independent of treatment allocation. The hospital pharmacy prepared morphine hydrochloride for intrathecal use as a sterile preservative-free 1 mg/mL solution supplied in 2-mL ampoules. For administration, a 0.1-mL volume, corresponding to 0.1 mg of morphine, was drawn using a 1-mL syringe.

Immediately after spinal puncture, patients were given ondansetron 4 mg IV and dexamethasone 4 mg IV, to prevent pruritus and PONV.^19,21 Upon confirmation of effective SA, light periprocedural sedation was provided as necessary using low doses of IV ketamine, midazolam, or propofol via target-controlled infusion (TCI), or short-acting IV opioids (alfentanil or remifentanil). Sedation depth was routinely monitored using the bispectral index (BIS), aiming to maintain values between 70–100. Although procedural sedation was not included among the predefined study variables, potential perioperative adverse events such as excessive sedation, respiratory depression, hypotension, or bradycardia were monitored and recorded in the case report forms (CRFs).

Outcomes

The primary outcome was postoperative pain intensity, which was measured using the NRS, at rest (every 2 hours) and during repositioning, over the first 24 hours. Secondary outcomes were the time to first analgesic administration (hours); number of non-opioid analgesics used (none, monotherapy, or dual-therapy); total opioid consumption in morphine milligram equivalents (MME); and incidence and severity of adverse effects, including respiratory depression (respiratory rate < 10/min, tidal volume < 4 mL/kg, SpO₂ < 90%, or PvO₂ < 4.5 kPa), bradycardia (HR < 50/min), hypotension (SBP < 90 mmHg, MAP < 60 mmHg, or a 20–30% drop from baseline), pruritus, and PONV. The classification of adverse effects severity is provided in the Results section.

Postoperative Management

Post-Operative Follow-Up

Postoperatively, all patients were transferred to the ICU and monitored according to institutional protocols for a minimum of 24 hours. While in the ICU, the following parameters were recorded: ECG, NIBP, HR, RR, SpO₂, and Richmond Agitation-Sedation Scale (RASS). Additionally, postoperative pain was measured using the NRS (0–10), along with the administered doses of both opioid and non-opioid analgesics. Pain intensity was self-reported by patients. Clinical staff were not involved in the assessment itself; their role was solely recording the reported values in the routine medical documentation and in the CRF. Patients were not informed of their treatment allocation, thereby minimizing the risk of observer bias for this subjective endpoint. Analgesia was provided based on reported pain intensity, with consistent application in both study groups. Every 2 hours, patients were asked about pruritus and PONV. At 24 hours after ICU admission, the investigator completed the CRF, documenting all primary and secondary parameters. Cross-checking against routine medical documentation was performed. Subsequent patient care was provided following standard institutional practice, guided by the individual clinical course.

Pain Management

Pain was managed following established institutional protocols, which complied with the 2022 Guideline for the Treatment of Acute Pain issued by the Czech Society of Anaesthesiology, Resuscitation, and Intensive Care Medicine (CSARIM).²² Throughout the 24-hour monitoring period, the first-line therapy included non-opioid agents such as paracetamol 1000 mg IV, metamizole 2500 mg IV, or diclofenac 75 mg IV. If the analgesic effect of non-opioid agents was insufficient, patients could be given opioid analgesics (tramadol, piritramide, or morphine), either via patient-controlled analgesia (PCA), continuous infusion, or IV boluses. The selection of specific agents from the above-mentioned medications and the method of their administration were at the discretion of the treating physician, based on the patient’s medical history and current clinical condition. Pain intensity was routinely assessed at least every 2 hours by clinical staff using patient self-report on a numeric rating scale (NRS, 0–10). Analgesics were given only in response to pain intensity, and not prophylactically. Analgesia was administered according to a stepwise protocol based on NRS pain thresholds, with non-opioid agents indicated for NRS ≥ 3/10 and opioids introduced at NRS ≥ 5/10. When possible, a combination of two non-opioid agents was preferred before escalation to opioids.

Adverse Effects Management

If respiratory depression occurred, oxygen supplementation was initiated to maintain adequate oxygenation and prevent hypoxia. Severe respiratory depression was defined as hypoxia or hypercapnia requiring naloxone administration.²³ Management consisted of naloxone 0.1 mg IV, administered every 2 minutes until recovery, followed by continuous infusion when necessary. Hypotension was managed with boluses of balanced crystalloid solution and, when required, an infusion of norepinephrine. Bradycardia was managed following the 2021 European Resuscitation Council guidelines,²⁴ with atropine 0.5 mg IV as first-line therapy. Pruritus was treated following a stepwise approach, with oral promethazine 25 mg as first-line therapy, ondansetron 8 mg IV as second-line treatment, and naloxone 0.1 mg IV as third-line therapy. PONV was initially managed with ondansetron 4 mg IV and dexamethasone 4 mg IV; if symptoms persisted, further treatment options included metoclopramide 10 mg IV, propofol 20 mg IV, or midazolam 1 mg IV.

Statistical Methods

Sample Size Calculation

It was initially estimated that a minimum of 50 patients would be sufficient to detect a difference between groups, based on prior studies and clinical experience.^15,25,26 After one year, when 42 patients had been enrolled, a blinded interim sample size re-estimation was conducted. The re-estimation was based solely on the observed variability in the dataset, using the highest observed standard deviation (SD = 1.78). Assuming a minimal clinically important difference of 1.27 points on the NRS (0–10), as reported by Kendrick et al,²⁷ and the current allocation ratio of 2:3, the recalculated sample size was 68 participants (27 in the control group and 41 in the intervention group). Thus, allowing for a 20% dropout rate, the final target sample size was set at 82 participants.

Statistical Analysis

All analyses were performed using R software.²⁸ A p value of 0.05 was considered to indicate statistical significance. Continuous and ordinal variables were summarized as medians with first and third quartile (Q1–Q3), and compared using the Mann–Whitney U-test. Welch’s t-test was initially considered due to potential variance heterogeneity; however, the Mann–Whitney test was chosen to avoid reliance on distributional assumptions altogether. Categorical variables were described as count (%), and analyzed using Fisher’s exact test.

Longitudinal changes in NRS scores over time were analyzed using a linear mixed-effects model fitted by restricted maximum likelihood (REML). NRS scores were modeled as a function of time, group, and their interaction (time × group). Random intercepts for participants were included to account for repeated measurements within individuals, and an additional random intercept for allocation week was incorporated to account for potential clustering related to the allocation procedure.

Incomplete data were planned to be excluded from the analysis. However, there were no missing data, as all parameters were recorded prospectively and cross-checked against routine medical documentation. The Holm method was used post hoc to adjust p values for multiple comparisons. The anonymized study dataset is provided as Supplementary Table 1.

Results

Both groups were comparable, and no significant differences were observed. Table 1 summarizes the baseline characteristics.

Table 1 Baseline Characteristics

The ITM group had a lower NRS score at rest and during repositioning within 24 hours after surgery. Regarding patients’ mean postoperative NRS scores at rest, the median was 0.38 (IQR 0.00–0.85) in the ITM group, and 1.73 (IQR 1.23–1.85) in the control group (adjusted p < 0.001). Regarding the mean postoperative NRS scores during repositioning, the median was 1.00 (IQR 0.00–1.77) in the ITM group, and 2.50 (IQR 1.89–3.44) in the control group (adjusted p < 0.001) (Figure 2). The median time to first analgesic administration was 16.5 (IQR 8.75–24.00) hours in the ITM group, and 7.00 (IQR 4.62–9.88) in the control group (adjusted p = 0.001). Postoperative analgesia was not required in 29.8% of patients in the ITM group, and 2.9% of patients in the control group (adjusted p = 0.001). Opioid consumption did not significantly differ between groups (adjusted p > 0.168) (Table 2).

Figure 2 Patients’ mean NRS pain scores at rest and during repositioning within 24 hours after surgery in the intervention and control groups. Box plots represent the median, first and third quartiles (Q1, Q3), and outliers of these means. The Holm method was used to adjust the p values. Abbreviation: NRS, numeric rating scale.

Table 2 Comparison of Pain Intensity Measured by 11-Point Numeric Rating Scale (NRS) and Analgesic Consumption Within 24 Hours Postoperatively, Between the Intervention and Control Groups

Longitudinal analysis of repeated pain measurements using a linear mixed-effects model demonstrated a significant effect of time and group. NRS scores increased by approximately 0.05 points per hour (p < 0.001), while patients in the treatment group had significantly lower NRS scores compared to the control group (p < 0.001). The interaction between time and treatment group was not statistically significant (p = 0.55), indicating that the rate of change in NRS scores over time did not differ between groups. Therefore, the trajectories over time appeared to be parallel, with similar increases in NRS scores across time (Figure 3). Random effects indicated moderate variability in baseline NRS scores between participants, whereas variability attributable to week was minimal, suggesting a limited clustering effect related to the allocation procedure.

Figure 3 Visual representation of repeated NRS scores at rest within 24 hours after surgery in the intervention and control groups. Abbreviation: NRS, numeric rating scale.

The incidence and severity of adverse effects did not significantly differ between the groups. Respiratory depression was observed in 21 patients (44.7%) in the ITM group, and 13 (38.2%) in the control group (adjusted p = 1.0). No cases of severe respiratory depression occurred. Hypotension was observed in 10 patients (21.3%) in the ITM group, and 4 (11.8%) in the control group (adjusted p = 1.0). Bradycardia was detected in 4 patients (8.8%) in the ITM group, and 1 (2.9%) in the control group (adjusted p = 1.0). PONV was observed in 8 patients (17.0%) in the ITM group, and 2 (5.9%) in the control group (adjusted p = 1.0). Pruritus occurred in 1 patient (2.1%) in the ITM group, and none in the control group (adjusted p = 1.0). Table 3 presents detailed results regarding the incidence and severity of adverse effects.

Table 3 Incidence and Severity of Adverse Effects Within 24 Hours After Surgery in the Intervention and Control Groups

Discussion

The present results suggest that compared to standard multimodal analgesia, 100 µg ITM may provide more effective analgesia in elderly patients during the first 24 hours after HF surgery. Analysis of the patients’ mean postoperative NRS scores at rest and during repositioning revealed a significantly lower median in the ITM group. The between-group difference on the NRS exceeded 1.27 points, indicating a clinically significant difference according to Kendrick et al²⁷ Furthermore, the ITM group showed a prolonged time to first analgesic administration, and reduced consumption of non-opioid analgesics. Opioid consumption did not significantly differ between groups and remained very low overall.

These findings are consistent with those reported by Murphy et al,¹⁴ who demonstrated that 100 µg ITM provides effective analgesia with an acceptable safety profile among elderly patients undergoing elective hip arthroplasty. Correspondingly, a recent study compared 100 µg ITM with sufentanil PCIA in elderly patients with HF, and reported superior pain control and reduced opioid requirements in the ITM group.¹⁵ Similarly, Yamashita et al²⁹ demonstrated that 50 µg ITM provides good postoperative analgesia without a higher incidence of side effects after bipolar hip prosthesis in patients aged 85 years or more.

In contrast, Abou-Setta et al⁶ performed a systematic review and found insufficient strength of evidence for both multimodal analgesia and spinal opioids in HF management. They reported moderate-level evidence for (PNBs).⁶ According to PROSPECT guideline for pain management after hip fracture repair surgery, single shot fascia iliaca compartment block or femoral nerve block are recommended. They should be applied in the pre-operative setting, to improve patients’ comfort prior to surgery. Other techniques such as lateral femoral cutaneous nerve block and pericapsular nerve group (PENG) block, offer the additional advantage of motor-sparing effects, however they are not recommended due to limited procedure-specific evidence.¹⁰ Since peripheral nerve blocks are becoming increasingly used for pain management, a prospective trial comparing their analgesic efficacy with intrathecal morphine would be valuable to further define the optimal analgesic strategy for HF surgery.²⁹

The most feared adverse effect of ITM is delayed respiratory depression, particularly in elderly patients. In the general population, the reported incidence is <1% with low-dose ITM, and varies according to dose and definition. In our elderly patient cohort, we observed a relatively high incidence of delayed respiratory depression in both groups (44.7% vs. 38.2%). The incidence did not significantly differ between groups. No patient in either group developed severe respiratory depression.

The definition of respiratory depression and the number of monitored parameters vary considerably across studies, which in turn leads to differences in the reported incidence.^12,29 In our study, we used a broad definition based on multiple monitoring parameters – respiratory rate, tidal volume, SpO₂, PvO₂, and the need for naloxone administration (as described in Postoperative Management – Outcomes). This may have led to the detection of subtle physiological changes that do not necessarily result in clinically significant events, potentially overestimating the incidence. The high rates of respiratory depression may have also reflected residual intraoperative sedation. Although no perioperative adverse events were observed, intraoperative respiratory depression may have been masked by preemptive oxygen supplementation when sedation was administered. This could have contributed to the need for oxygen supplementation after admission to the ICU. This finding may reflect the increased vulnerability of the elderly study population to respiratory depression. Future studies should incorporate time-based assessment of respiratory parameters, including the duration of abnormalities, and consider refining the thresholds used to define respiratory depression and guide treatment initiation.

The most common adverse effect of ITM is opioid-induced pruritus, which occurs with an incidence of 30–60% among orthopedic patients.³⁰ The risk significantly increases with doses above 100 µg, but can be minimized by using low-dose ITM or by prophylaxis with a 5-HT3 antagonist.³¹ In our study, pruritus occurred in one patient in the ITM group, who did not require treatment, and was not observed in any patient in the control group. No significant difference was found between the groups. These findings may be attributable to the combined use of a low-dose regimen and application of pharmacological prophylaxis.

Although patients over 65 years of age are more susceptible to opioid-related adverse events and have an increased risk of respiratory depression,³ intrathecal opioids are still recommended as part of multimodal analgesia because they reduce the need for systemic opioids.³⁰ Low-dose regimens enable their safe use.¹⁸ Considering that HF repair is associated with moderate-to-severe pain and that the consequences of undertreated pain in elderly patients are substantial, the search for improved postoperative analgesic strategies is clinically important.¹⁰ Because postoperative pain intensity may depend on fracture type, surgical procedure, and institution-specific techniques, further multicenter studies focusing on specific surgical subgroups are needed. No serious or unexpected adverse events occurred during the study, and the incidence and severity of adverse effects were comparable between groups. However, reported incidence of respiratory depression was unusually high in both groups.

This trial provides important evidence regarding the safety and efficacy of low-dose intrathecal morphine in elderly patients and suggests that this approach may represent a promising option for postoperative analgesia after HF surgery.

Limitations

Several limitations should be acknowledged. This study was conducted at a single center with a small sample size, which may limit the external validity of the results, since perioperative care can vary across institutions. The weekly allocation strategy may have been a potential source of selection bias, although additional analysis suggested a minimal clustering effect, indicating that its impact on the results was likely limited. The clinical staff were aware of the treatment allocation due to the single-blind design, which might have influenced the evaluation of subjective endpoints, such as pain or certain adverse effects. The use of a broad definition of respiratory depression based on multiple monitoring parameters may have led to the detection of clinically insignificant events, potentially resulting in an overestimation of its incidence in this study population. A uniform dose of ITM was administered without adjustment for BMI, which may have impacted both analgesic efficacy and the risk profile. Finally, a longer postoperative follow-up period than 24 hours would be more appropriate to detect delayed respiratory depression, given the increased susceptibility of this patient population to adverse effects and the existing evidence gap in elderly patients with hip fracture.

Conclusion

Compared with standard multimodal analgesia, intrathecal morphine administration (100 µg) provides superior postoperative analgesia in elderly patients undergoing single spinal anesthesia for hip surgery, without increasing the incidence or severity of adverse effects. These results suggest that the administration of 100 µg ITM appears to be a safe approach, even in the elderly population studied. A relatively high incidence of delayed respiratory depression was observed in both groups, which may be attributable to sensitive detection criteria, as well as the population’s increased vulnerability.