Assessing Pain Control and Satisfaction in Spine Surgery Amidst COVID-19 Staff Shortages: A Retrospective Study on Patient-Controlled Analgesia

Abstract

Background:

Effective pain management is critical for recovery after spine surgery, as inadequate control can delay mobilization, prolong hospitalization, and raise opioid consumption.^1–4 Among the strategies available, patient-controlled analgesia (PCA) enables patients to self-administer intravenous opioids within preset safety limits. Inconsistencies in opioid use, pain scores, and length of stay suggest that the effectiveness of PCA may be shaped by contextual and institutional variables. One example being the significant strain the COVID-19 pandemic placed on hospital staffing and perioperative workflows, potentially undermining the effectiveness of provider-administered analge-sia.^9–10 Although postoperative pain management has been previously studied, few investigations have examined how pandemic-related staffing shortages may have impacted the comparative effectiveness of PCA versus non-PCA strategies.

Methods:

This retrospective cohort study included 5,528 adult patients who underwent one of four spine procedures between January 2018 and December 2023. Patients were stratified by analgesia type (PCA vs. non-PCA) and time period (pre-COVID vs. post-COVID). Primary outcomes included total opioid consumption (measured in morphine milliequivalents, MME) and average Visual Analog Scale (VAS) pain scores in the first 24 hours postoperatively. Multivariable regression adjusted for demographics including age, sex, obesity, and smoking.

Results:

PCA use was associated with a 52.1% reduction in opioid consumption compared to non-PCA (p < 0.001). This effect was most pronounced post-COVID (74.2% reduction), though not statistically significant. Despite reduced opioid use, PCA patients reported slightly higher VAS scores (+0.814, p = 0.0013). Subgroup analyses revealed higher opioid use and pain scores among smokers and obese patients, while male sex predicted lower pain scores. PCA was particularly beneficial in more invasive procedures, and its expanded use in less complex surgeries post-COVID did not increase pain scores.

Conclusion:

PCA significantly reduced opioid use following spine surgery, though pain scores were modestly higher. During staffing shortages, PCA may offer operational and clinical advantages. Patient-specific risk factors should guide individualized pain protocols.

Level of Evidence:

III

Introduction

Effective pain management is critical for recovery after spine surgery, as inadequate control can delay mobilization, prolong hospitalization, and raise opioid consump-tion.^1–4 Among the strategies available, patient-controlled analgesia (PCA) enables patients to self-administer intravenous opioids within preset safety limits. This strategy has become widely used due to its ability to deliver timely pain relief.^5–7 However, while PCA promotes patient autonomy and may reduce delays in pain control, studies have shown mixed results: some report lower opioid use, while others show no significant differences in pain scores or hospital length of stay when compared to provider-administered opioids.^7–9 These inconsistencies suggest that the effectiveness of PCA may be shaped by contextual and institutional variables.

One contextual factor to consider is the significant strain the COVID-19 pandemic placed on hospital staffing and perioperative workflows, potentially undermining the effectiveness of provider-administered analgesia.^9–10 In such extreme staffing shortages, PCA may offer clinical and operational advantages by minimizing the need for frequent nurse involvement while ensuring consistent pain relief. Although postoperative pain management has been previously studied, few investigations have examined how pandemic-related staffing shortages may have impacted the comparative effectiveness of PCA versus non-PCA strategies. This study addresses that gap by retrospectively comparing opioid consumption and pain scores in PCA and non-PCA patients undergoing spine surgery, before and during/after the CO VID-19 pandemic, to identify approaches that optimize patient outcomes and system efficiency.

Methods:

Study Design

This was a retrospective cohort study conducted at a single tertiary academic institution, evaluating the impact of PCA compared with non-PCA pain control in adult patients undergoing spine surgery. To encompass the time before and during the COVID-19 pandemic, the study period spanned from January 1, 2018, through December 8, 2023. All data collection and procedures were conducted in accordance with IRB standards and guidelines.

Patients

Inclusion criteria for this study consisted of adults >18 years old who underwent one of four predefined spine surgeries: 1) anterior cervical discectomy and/or fusion, 2) posterior cervical fusion, 3) posterior thoracic or lumbar fusion, or 4) laminectomy/discectomy. Eligible participants also needed to have received postoperative pain management via either PCA or non-PCA methods. Exclusion criteria included history of opioid allergies, history of substance use disorder, or patients who were under the age of 18 at the time of surgery.

Interventions

Patients were stratified into PCA or non-PCA groups based on the type of postoperative pain control administered. The PCA group received self-administered intravenous opioids via a programmable pump without a basal rate. The non-PCA group received standard healthcare professional-administered intravenous or oral opioid dosing, dependent on provider and institutional protocols. All patients had additional access to non-opioid supplemental pain control such as acetaminophen, gabapentin, muscle-relaxers and non-steroidal anti-inflammatory drugs unless contraindicated. Opioid dosing in both the PCA and non-PCA groups were converted to morphine milliequivalents (MME) for appropriate and reliable comparison.

Data Collection

Data was collected through chart review using the electronic medical record system. This included patient demographics (age, gender, BMI, smoking status), surgery type, pain control methods (PCA vs non-PCA). The primary outcomes were average daily opioid consumption (in MME) during the first 24 hours postopera-tively and pain control as assessed by the Visual Analog Scale (VAS) as an average. The presence of healthcare worker shortages during the COVID-19 pandemic was recorded as a binary variable, with surgeries performed after March 2020 marked as “post-COVID,” indicating that those cases were COVID-affected. Data was stratified by procedure type and pain management method to examine trends across subgroups and time periods.

Statistical Methods

Patient demographics, including obesity, smoking status, sex, and age, were summarized and stratified according to surgery type and analgesia method. The primary outcomes included daily total MME and daily average VAS scores; these were compared across groups using multiple linear regression, adjusting for patient demographics as potential confounders. Interactions between confounders were considered but not included in the final model due to lack of statistical significance. To correct for skewness, daily total MME was log-transformed prior to analysis. A significance threshold of p < 0.05 was used for all statistical tests.

Results:

Baseline Characteristics

A total of 5,528 adult spine surgery patients from January 2018 to December 2023 were included and stratified by pain management strategy (PCA vs. non-PCA) and era (pre-COVID vs. post-COVID). Surgical procedures included anterior cervical discectomy and fusion, posterior cervical fusion, posterior thoracic/lumbar fusion, and laminectomy/discectomy. Notably, PCA was less commonly used across all procedure types (Figure 1a-1d).

Figure 1A.

Percent of Patients Obese by Procedure and PCA Status.

Figure 1D.

Mean Patient Age by Procedure and PCA Status.

Figure 1B.

Percentage of Smokers by Procedure and PCA Status.

Figure 1C.

Percent of Male Patients by Procedure and PCA Status.

Opioid Consumption (MME)

PCA use was associated with reduced opioid consumption in the first 24 hours postoperatively. In adjusted analysis, PCA was linked to a 52.1% reduction in MME compared to non-PCA (-0.737, p < 0.001). PreCOVID, PCA patients had lower opioid requirements across procedures: ACDF (median 40 MME PCA vs. 92 MME non-PCA), posterior cervical fusion (median 76 MME vs. 88 MME), TLIF (median 56 MME vs. 114.5 MME), and laminectomy/discectomy (median 68 MME vs. 80 MME). Post-COVID, median opioid consumption remained lower for PCA patients: posterior cervical fusion (median 40 MME vs. 62 MME), TLIF (median 46 MME vs. 70 MME), and laminectomy/discectomy (median 28 MME vs. 52 MME). Overall, by procedure type, PCA patients consumed less than non-PCA patients: ACDF (median 40 MME vs. 70 MME), posterior cervical fusion (median 47.8 MME vs. 72.0 MME), TLIF (median 48 MME vs. 100 MME), and laminectomy/discectomy (median 36.8 MME vs. 64.0 MME) (Figures 2-3).

Figure 2.

Median Opioid Use (MME) by Procedure, PCA Status, and Time Period.

Figure 3.

Median Opioid Use (MME) by Procedure Type and PCA Status Only.

Posterior thoracic/lumbar fusion patients receiving PCA post-COVID showed a trend toward reduced opioid use compared to non-PCA, but this subgroup effect was not statistically significant (p = 0.334). For posterior cervical fusion procedures, PCA patients had lower opioid requirements post-COVID (median 40 MME) compared to non-PCA patients (median 62 MME). In laminectomy/discectomy cases, PCA use increased, and mean pain scores remained stable or improved slightly with expanded PCA use. Across procedures, PCA use increased post-COVID in posterior cervical fusion (234 vs. 132 pre-COVID), TLIF (1270 vs. 621), and laminectomy/ discectomy (51 vs. 30), while ACDF had few PCA cases post-COVID (3 vs. 79 pre-COVID). Non-PCA counts were relatively stable or increased for ACDF (726 vs. 582), TLIF (231 vs. 368), posterior cervical fusion (87 vs. 89), and laminectomy/discectomy (522 vs. 503) (Figure 4).

Figure 4.

PCA and Non-PCA Patient Counts by Procedure Type and Time Period.

Smokers consumed 21% more opioids than nonsmokers (p = 0.194, p < 0.001). Age was associated with a slight decrease in opioid use (p = -0.011), whereas male sex (p = 0.011) and obesity (p = 0.052) were not significantly associated with differences in opioid consumption (Figure 6b).

Figure 6B.

Adjusted Effect of Patient Factors on Log MME.

Pain Control (VAS Scores)

During the first 24 hours postoperatively, PCA patients reported slightly higher pain scores compared with non-PCA patients. Pre-COVID, mean VAS scores (PCA vs non-PCA) were 5.1 versus 4.0 for ACDF, 5.7 versus 4.0 for posterior cervical fusion, 5.5 versus 5.2 for TLIF, and 5.2 versus 4.1 for laminectomy/discectomy. Post-COVID, mean VAS scores remained higher for PCA patients: 6.2 versus 4.4 for ACDF, 5.8 versus 5.0 for posterior cervical fusion, 5.7 versus 5.0 for TLIF, and 4.6 versus 3.9 for laminectomy/discectomy (Figure 5). Across all procedures, the overall mean difference in VAS between PCA and non-PCA patients was 0.814 points (p = 0.0013), with differences persisting in the pre-COVID (+0.965, p = 0.0003) and post-COVID (+1.199) eras, although the change between eras was not statistically significant (A = +0.234, p = 0.823).

Figure 5.

Mean VAS Pain Scores by Procedure, PCA Status, and Time Period.

Estimated VAS pain scores varied across patient traits (Figure 7). Patients who used PCA reported the highest pain scores (VAS = 6.72), followed by smokers (VAS = 6.41), obese patients (VAS = 6.01), and female patients (VAS = 5.91). Male patients reported lower pain scores (VAS = 5.44), and pain scores decreased modestly with age (VAS = 4.00 at age 60 vs 4.63 at age 40). These differences reflect regression-estimated effects: PCA use was associated with the largest increase in pain (p = 0.814), followed by smoking (p = 0.502) and obesity (p = 0.106). In contrast, male sex (p = -0.467) and increasing age (p = -0.032 per year) were associated with lower pain scores. These effects were generally consistent across pre- and post-COVID periods, although some differences were attenuated post-COVID (Figure 6a).

Figure 7.

VAS Score Comparison Across Binary Traits.

Figure 6A.

Adjusted Effect of Patient Factors on VAS Pain Scores.

Discussion

In this study, PCA was associated with a significant reduction in opioid consumption during the first 24 hours following spine surgery, with an overall 52.1% decrease in MME compared to non-PCA patients. This opioid-sparing benefit was even more pronounced during the post-COVID era, which may reflect the relative inefficiency of non-PCA methods when staff shortages delayed timely analgesia administration. PCA systems allow patients to manage their pain proactively, providing autonomy that is particularly beneficial during periods of limited nursing availability (Figures 2-4).

Despite the reduction in opioid use, PCA was associated with a mild increase in pain scores, with an overall mean difference of 0.814 VAS points. This trend was consistent across pre- and post-COVID periods, suggesting that while PCA effectively reduces opioid exposure, it does not necessarily improve subjective pain perception. These findings mirror prior work indicating that PCA enhances patient autonomy and reduces opioid use without substantially improving pain scores or satisfac-tion.^7–8 The slightly higher VAS scores observed may reflect cautious PCA settings, underdosing, differences in patient expectations, or timing of assessments.

Procedure-specific analyses revealed nuanced effects. PCA was particularly effective for more invasive surgeries, such as posterior thoracic or lumbar fusion (TLIF), where opioid use was substantially lower in PCA patients compared with non-PCA patients (pre-COVID: 56 vs. 114.5 MME; post-COVID: 46 vs. 70 MME). In less complex procedures, like laminectomy/discectomy, PCA use increased post-COVID (51 vs. 30 pre-COVID), but pain scores remained similar between PCA and non-PCA patients (pre-COVID VAS: 5.2 vs. 4.1; post-COVID VAS: 4.6 vs. 3.9), suggesting that PCA can be safely expanded to these procedures without increasing patient-reported pain (Figure 4). These findings highlight the value of tailoring PCA use to both procedure invasiveness and clinical context.

Patient traits further influenced opioid consumption and pain perception. Smokers used more opioids than non-smokers (21% higher MME; p = 0.194) and reported higher pain scores (VAS p = 0.502), likely reflecting altered pain thresholds and opioid tolerance. Obese patients had slightly higher pain (VAS p = 0.106), potentially due to pharmacokinetic differences and increased mechanical strain during the postoperative period. Male sex and older age were associated with lower pain scores (VAS p = -0.467 and -0.032 per year, respectively) and modestly reduced opioid use (male p = 0.011; age p = -0.011). These patterns, illustrated in Figures 6a-6b and 7, should be considered when planning postoperative pain management, regardless of analgesia delivery method.

Overall, PCA represents an effective component of multimodal analgesia, reducing opioid burden while providing patients with autonomous access to pain control. However, this benefit comes with a modest increase in pain scores, which should be considered for patients at higher baseline risk for postoperative pain such as smokers and those with obesity. These findings emphasize the importance of individualized pain management strategies that account for procedure type, patient characteristics, and contextual factors such as staffing availability.

Conclusion:

While PCA effectively lowered opioid use following spine surgery, its use was associated with mildly elevated pain scores. These findings suggest that PCA offers operational and clinical advantages - particularly during staffing shortages - but should be deployed thoughtfully, considering patient-specific risk factors.

The retrospective design limits causal inference, and unmeasured confounders (e.g., anxiety, individual pain sensitivity) may influence both analgesia type and outcomes. Variation in non-PCA delivery methods (e.g., IV vs. oral opioids) introduces additional complexity. Moreover, key social determinants of health - including race, primary language, and insurance status - were not assessed but may impact access to PCA or pain outcomes.

Future prospective research should incorporate standardized protocols, long-term patient-reported outcomes, and education around PCA expectations. Evaluating disparities in access and tailoring PCA settings to individual risk will be essential as healthcare systems prepare for continued staffing challenges. PCA may be a scalable, patient-centered strategy - particularly when paired with non-opioid adjuncts and individualized care planning.