Pain After Lower Extremity Fracture Surgery: What Is Normal and How Is It Impacted by Individual-specific and Caregiver-specific Psychosocial Factors?

Abstract

Background

This pilot study characterized normative pain and analgesic consumption following lower-extremity fracture surgery in children. Additionally, we sought to determine how various caregiver- and patient-specific psychosocial factors affect pain, opioid use, and health-related quality-of-life (HRQOL). We hypothesized that the majority of children following lower extremity fracture surgery would not need prolonged opioids, and that higher preoperative anxiety levels would translate into higher postoperative pain and analgesic needs.

Methods

This was a prospective cohort study of children aged 5–17 years old undergoing operative fixation of an isolated lower extremity fracture at a single pediatric trauma center. Baseline surveys were administered preoperatively assessing anxiety, pain catastrophizing, and pain self-efficacy. Daily pain scores and analgesic consumption were collected on postoperative days (POD) 1–7, 10, 14, and 21. HRQOL was assessed via PROMIS Global Health, and functional recovery was assessed via PROMIS Mobility. Clinical data were abstracted from the medical record. Descriptive and bivariate statistics were performed.

Results

A total of 63 patients with a mean age of 13.0 ± 2.6 years were included. The most common fracture locations were the ankle (49%), tibia (27%), and femur (17%). Mean daily pain scores peaked at 4.5/10 on POD1 and steadily decreased thereafter. By POD3, 62% of patients had mild pain or less (score ≤3/10), and 53% were no longer taking any opioids. The majority of caregivers (55%) and patients (61%) reported pain-related anxiety of ≥6/10 prior to surgery; 76% of caregivers and 50% of patients were at least moderately confident in their ability to adequately manage pain. Neither caregiver nor patient pain-related anxiety or self-efficacy was significantly related to postoperative pain scores, satisfaction with pain management, or opioid consumption. Caregiver pain catastrophizing and generalized anxiety also did not affect patient outcomes. One week after surgery, anxious children had lower average pain scores (1.2 vs 2.7; P = .005) and were less likely to still be using nonopioid analgesics (47% vs 88%, P = .013).

Conclusions

This pilot study contributes preliminary data for postoperative pain, analgesic consumption, and functional recovery following operative treatment of lower extremity fractures that can be used to guide prescribing practices and preoperative expectation management/family education. More work is needed to validate and build upon these findings before they can be safely extrapolated to other clinical scenarios and age groups.

Key Concepts

• (1)In this cohort of children following lower extremity fracture surgery of all types, mean daily pain scores peaked at 4.5 on postoperative day one and steadily decreased thereafter.
• (2)By POD3, most patients had mild pain or less and the majority were no longer taking any opioid analgesics.
• (3)The majority of children continue to take nonopioid analgesics (acetaminophen and/or ibuprofen) for a week (or longer) after lower extremity fracture surgery.
• (4)In this cohort, high-anxiety children tended to report lower pain and take fewer opioids after lower extremity fracture surgery.
• (5)Although baseline pain-related anxiety was not significantly associated with postoperative outcomes, it remained a significant concern for many families and children, suggesting further work is necessary to determine if, and to what extent, targeted interventions might be helpful.

Level of Evidence

II, prospective cohort study

Introduction

Lower extremity fractures are common in children. While many injuries can be successfully treated with nonoperative management, others require surgical intervention, oftentimes resulting in substantial postoperative pain. A child's unique postoperative pain experience involves a complex interplay between various biological, psychological and socio-environmental factors [[1], [2], [3], [4], [5]]. In particular, previous work has demonstrated the importance of baseline psychosocial variables such as anxiety, pain catastrophizing, and pain self-efficacy on clinical outcomes and pain intensity following major pediatric orthopaedic surgery as well as other large nonorthopaedic procedures [[6], [7], [8], [9]]. Parents also have an important mediating effect on their child's pain experience [[10], [11], [12], [13], [14]].

Despite recent emphasis on prescribing practices and minimizing opioid use in children following orthopaedic surgery [[15], [16], [17]], few studies report normative data on pain and medication use following lower extremity orthopaedic trauma procedures in children and adolescents. Furthermore, there have been no studies to date exploring the importance of psychosocial factors following smaller, less morbid orthopaedic procedures in children.

The purpose of this pilot study is twofold. First, to help guide preoperative family education, we aim to present normative data for postoperative pain and analgesic use following lower extremity fracture surgery in children. Second, we sought to identify preoperative psychosocial risk factors for postoperative pain intensity and health-related quality-of-life following pediatric musculoskeletal trauma. We hypothesized that the majority of children following lower extremity fracture surgery would not need prolonged opioids, and that higher preoperative anxiety levels would translate into higher postoperative pain and analgesic needs.

Materials and methods

This was a prospective cohort study at a single level-I pediatric trauma center. Institutional review board approval was obtained. Children aged 5–17 years old undergoing operative fixation of an isolated lower extremity fracture between November 2021 and February 2023 were screened. Exclusion criteria included: 1) history of chronic opioid use or pain; 2) open fractures; 3) vascular injuries; 4) metabolic or genetic bone disease; 5) underlying neuromuscular, syndromic, or developmental diagnosis; 6) non-English speaking; 7) no access to a mobile phone; 8) polytrauma of ≥2 extremities or organ systems; 9) >2 weeks between injury and surgery. We included both cold trauma (fractures treated electively during a separate admission following initial outpatient evaluation) and hot trauma (fractures treated operatively during the same admission as the initial presentation to the emergency department).

Following informed consent, enrolled children, and caregivers completed the baseline survey together preoperatively or by parent-proxy within 24 h of surgery. At baseline, generalized anxiety was assessed in children (PROMIS Anxiety, short form 8) and caregivers (Generalized Anxiety Disorder 7 [GAD7] [18]). Pain catastrophizing in parents was assessed via the Pain Catastrophizing Scale – Parental Version (PCS–P) [13]. Finally, single-item measures for both pain-related anxiety and pain confidence (pain self-efficacy) were assessed in children and caregivers [19]. Specifically, anxiety was assessed using the prompt “how worried or concerned are you (or your child) about postoperative pain” while pain confidence (self-efficacy) was assessed using the prompt “how confident are you (or your child) in your family's (or his/her) ability to manage the level of pain experienced after surgery.” Each was scored from 1 to 10.

Following completion of the baseline survey, all enrolled families received a text message link to questionnaires on postoperative days (POD) 1–7, 10, 14, and 21. The distribution of questionnaires was automated through REDCap (Research Electronic Data Capture; Vanderbilt University), using validated text message-based protocols. [[20], [21], [22], [23], [24]]^. Postoperative pain and analgesic consumption were assessed using a series of questions previously validated in pediatric supracondylar humerus fractures [20,21]. Questions included a Wong-Baker FACES Pain Rating Scale, the type(s) of analgesic (opioids, acetaminophen, and ibuprofen) used and number of doses of each taken in the last 24 h.

Postoperative outcomes were assessed through Patient Reported Outcome Measurement Information System (PROMIS) measures for functional recovery (PROMIS Mobility) and HRQOL (PROMIS Global Health) at various timepoints (Fig. 1). The family's overall satisfaction with pain management was assessed on a 5-point Likert scale (0 = Very Unsatisfied, 5 = Very Satisfied).

Figure 1.

Timing of postoperative questionnaires administered on each postoperative day (POD) following surgery.

Perioperative data were recorded retrospectively, including patient demographics, fracture type, and postoperative weight bearing. The following clinical assumptions were made to extrapolate for incomplete/missing data: 1) if analgesic consumption was marked as “No” two days in a row (with no further self-entered “Yes” at any time), then all missing days moving forward were marked “No”; 2) if analgesic consumption was marked as “Yes” on any given day, then all missing days before that timepoint were marked “Yes”; 3) if families reported “not sure” with respect to analgesic consumption, then they were assumed to still be taking that medication (“Yes”) for that given day; 4) if there was a self-entered pain score of “0” three days in a row, all missing days moving forward were marked as “0”; 5) if families missed a pain diary entry, but had the same score on self-entered values the days before and after, then the same score was entered for the missing day.

Results were analyzed using descriptive and bivariate statistical analysis as appropriate via SAS 9.4 software. All comparative analyses were 2-tailed with alpha set at 0.05. Patients were counted as “taking a medication” on a given day if they took a dose that day, or on any subsequent days in the future.

Results

A total of 106 patients were enrolled, of which 63 were eligible for inclusion in the final analysis (Fig. 2). These 63 patients had a mean age of 13.0 ± 2.6 years old and seventy percent (n = 44) were males. The most common fracture locations were the ankle (49%), tibia (27%), and femur (17%). All were treated with open reduction and either internal or percutaneous fixation.

Figure 2.

Flow diagram for patient enrollment, exclusion, and inclusion.

Mean daily pain scores peaked at 4.5/10 on POD1 and steadily decreased to <2/10 by POD10 (Fig. 3A). Most patients (32/55, 62%) had mild pain or less (≤3/10) by POD3, and only one patient (1/51, 2%) had severe pain (≥7/10) after POD7 (Fig. 4). Younger children (aged <10 years) tended to report less postoperative pain compared to early adolescents (11–14 years) and adolescents (15–17 years), but these differences did not reach statistical significance (Fig. 3B). Most caregivers were satisfied with their child's pain management on POD3 (37/52, 71%) and POD6 (40/46, 87%).

Figure 3.

Mean daily pain scores across all fracture types by postoperative day for A) all ages combined, and B) separated by age group.

Figure 4.

Pain categorization by postoperative day.

In total, 33% (21/63) of patients reported taking no opioids throughout the entire postoperative course, and the majority (28/53; 53%) were no longer taking opioids by POD3 (Fig. 5). No patients required opioids after POD10. Of those who did use opioids at any time (n = 42), the mean number of self-reported doses consumed was 5, and only 17% (7/42) took 10+ total doses. In contrast, nonopioid analgesic consumption was higher, with 73% of patients (35/48) taking these medications through POD7, and 14% of patients (6/42) still taking them up to 3 weeks postsurgery. There was no statistically significant relationship between age and opioid or nonopioid consumption (P > .05).

Figure 5.

Families self-report of continued opioid or nonopioid (e.g. acetaminophen or ibuprofen) consumption by postoperative day.

Prior to surgery, most caregivers (n = 34/61; 55%) and patients (n = 31/51; 61%) reported pain-related anxiety of 6 or higher on a scale of 1–10 using the single-item assessment of pain anxiety. As compared to their parents, children were more concerned about their ability to manage pain after surgery; 76% of caregivers (n = 48/63) were at least moderately confident in their family's ability to adequately manage the pain after surgery, but only 50% of patients (n = 25/50) were similarly confident (single-item assessment of pain confidence; Fig. 6). Neither caregiver nor patient preoperative pain-related anxiety or pain confidence were related to postoperative mean daily pain scores, satisfaction with pain management, or opioid/nonopioid analgesic consumption (all P > .05). Finally, there was no evidence of a relationship between caregiver pain anxiety/confidence and total postoperative opioid use after controlling for average postoperative pain (all P > .05).

Figure 6.

Baseline assessment of the family's pain-focused anxiety and self-efficacy prior to surgery.

Patients' mean HRQOL was equivalent to population norms on POD7 (PROMIS Global Health 49.4 ± 8.8) and POD14 (49.0 ± 8.2). By POD7, the majority of patients rated their physical health (38/49, 78%), mental health (34/49, 69%), and overall quality-of-life (41/49, 84%) as “very good” or “excellent.” Conversely, the patient's mobility was consistently decreased throughout the 3-week postoperative period, with a PROMIS mobility score of 25.2 ± 5.2 on POD10 and 27.2 ± 5.2 on POD21.

At baseline, 35% of parents met criteria for mild anxiety or more based upon the GAD7 diagnostic criteria (21/60 with score ≥5), while the mean parental catastrophizing score was 18 ± 11 using the Pain Catastrophizing Scale (PCS-P). Neither parental catastrophizing nor generalized anxiety on either of these measures was associated with postoperative pain or the proportion of children still taking any form of analgesic at any time point postoperatively (all P > .05). Looking at children, patients’ mean PROMIS Anxiety score before surgery was 51 ± 10, and 46% met criteria for anxiety (PROMIS score ≥50). Anxious children had lower average daily pain scores on POD7 (1.2 vs 2.7; P = .0054) and POD 10 (1.0 vs 2.1; P = .145). A higher proportion of these children also had pain characterized as mild or less (≤3/10) on POD7 [21/22 (95%) vs 19/27 (70%), P = .031). Compared with low-anxiety children, anxious children were less likely to still be taking nonopioid analgesic on POD5 [12/19 (63%) vs 23/25 (92%), P = .0270], POD7 [8/17 (47%) vs 21/24 (88%), P = .0127] and POD10 [2/14 (14%) vs 16/24 (67%), P = .0025). There was no relationship between anxiety and postoperative HRQOL (PROMIS Global Health) on POD7 or POD14 (all P > .05).

Discussion

An important, but often underappreciated, component of evidenced-based opioid prescribing practices is setting appropriate preoperative expectations. This study aimed to generate descriptive data for pain, analgesic consumption, and functional recovery following operative treatment of lower extremity fractures in children that providers can use to inform family discussions. In turn, our goal was to pinpoint opportunities for targeted, evidenced-based interventions that surgeons can preemptively use to reduce the risk of poorly managed postoperative pain and minimize functional impact.

We found children's postoperative pain and opioid consumption improved rapidly following lower extremity fracture surgery. Specifically, most patients were in minimal pain and no longer taking any opioids by POD3, and no patients reported opioid consumption after POD10. These findings built on previous work, including Nelson et al.’s report on closed reduction and percutaneous pinning for supracondylar humerus fractures that found pain and opioid usage decreased to clinically unimportant levels after POD3 [20]. Accounting for elevated pain associated with lower extremity fracture fixation, which is often open, our study corroborates their conclusion that most children recover quickly and uneventfully following fracture surgery. However, it is important to acknowledge a subset of patients in our cohort (16%) that continued to report moderate/severe pain at POD10 and beyond. Although this subgroup was a minority of patients, these findings demonstrate that persistent pain outside the norm should prompt consideration of additional clinical or psychosocial causes that may be contributing to a patient's recovery.

Nonopioid analgesics have been shown to be equally effective as opioids for pediatric upper extremity fractures [21]. For the lower extremity, we found that 73% of patients took nonopioid analgesics through POD7, and 14% of children were still taking them up to 3 weeks postsurgery. Such information is helpful for guiding prescribing practices and counseling families. Williams et al. reported on the implementation of a state-mandated opioid consent process and noted an increase in written prescriptions for nonopioid analgesics with this protocol [25]. In turn, they posit that families may be more likely to adhere to medication instructions for a prescribed medication, even if it is the same pill that can be obtained over-the-counter. Although we did not specifically monitor prescription patterns or continued nonopioid analgesic use in our patients suggest that specific instructions and prescriptions might be helpful for families.

Despite significant limitations in mobility through POD21, the patient's HRQOL remained high, with most patients rating their physical health, mental health, and overall quality-of-life as “very good” or “excellent.” Overall PROMIS global health scores were equivalent to population norms on both POD7 and POD14. Our finding of significant postoperative impaired physical mobility is consistent with prior work evaluating children's gait and physical activity after lower extremity surgery [[26], [27], [28]]. Taken together, these findings speak to the resilience of children, and the importance of proactively managing recovery expectations.

Postoperative pain is top-of-mind for most families undergoing pediatric orthopaedic surgery. Most caregivers and patients in this cohort had high pain-related anxiety (score ≥6/10) at baseline. Children feel particularly ill-equipped to manage pain, as only 50% of patients were at least moderately confident in their ability to manage pain. Although we did not find any relationship between caregiver pain catastrophizing or generalized anxiety and patient outcomes, we did find that anxious children (PROMIS Anxiety T-Score >50) had lower average pain scores (1.2 vs 2.7) and were less likely to still be using nonopioid analgesics (47% vs 88) one week after surgery. Previous work has suggested that pain catastrophizing is associated with pain intensity and pain interference in adult lower extremity trauma [29]. Additionally, in a meta-analysis looking at chronic post-surgical pain, Rabbitts and colleagues noted that preoperative pain intensity, child anxiety, child pain coping efficacy, and parental pain catastrophizing were predictive of postoperative pain [7]. Each of these factors are part of the biopsychosocial model of pain control [30]. Although such factors are well-documented to affect outcomes in major pediatric orthopaedic surgery, such as spinal fusion or hip preservation [[6], [7], [8], [9], [12], [14], [31], [32], [33]], [9,12,14,31] [[6], [7], [8], [9], [12], [14], [31], [32], [33]] to the authors' knowledge, there has been no work documenting predictive value in smaller, more common procedures such as fracture fixation. This differential impact of smaller/less invasive surgery may explain why our findings were counterintuitive (more anxiety equated to less analgesics). It's possible that in today's world of (appropriately) heightened concern regarding opioids, anxious children are more aware of the risks, and thus less likely to take them. Very anxious families may also go to greater lengths to prepare themselves, and their home environments, thus improving the postoperative experience and reducing the far-ranging impact surgery and limited postoperative mobility can have. While these strategies may not be enough for the rigors of major orthopaedic reconstructions, they may be sufficient for the lower analgesic demands of fracture surgery. Alternatively, these families may have assumed the worst, and then been surprised when recovery was easier than anticipated. However, it is important to note that each of these theories is purely hypothetical and requires additional investigation. While preoperative expectations and affective forecasting are undoubtedly powerful forces, we do not yet know enough to say exactly how each interacts with psychosocial differences across individuals to influence patient-reported outcomes.

Notably, more work is needed to validate these findings across additional cohorts. Just as our findings ran counter to previous conclusions drawn following major pediatric orthopaedic surgery, the findings described here are not conclusive, and it remains to be seen how various psychosocial factors may or may not contribute to different patient-reported outcomes in other clinical scenarios and procedure types. Caution should be exercised before these results are further extrapolated. However, at the very least, our study supports the conclusion that pain-anxiety and pain self-efficacy are substantial concerns for families, and in turn warrant perioperative counseling from surgeons. While the focus and content of such counseling will vary among different patients and procedures, such efforts might focus on setting expectations, counseling families on normative data (as presented here), and/or discussing strategies to manage pain (e.g. medications, teaching coping strategies, connecting to child life therapy/pain psychologists). Additionally, larger psychosocial interventions better prepare families to cope with surgery and pain may be valuable and well-received [19,[34], [35], [36]].

This study has several limitations. First, this was a single-center study with a response rate of 61%, which may have introduced selection bias as these patients may have had more pain than nonresponders. However, if anything, this would tend to overestimate pain following surgery; thus, counseling families based on our conclusions would still be valid. Additionally, as a single-center study, our practice may not accurately reflect the considerable variation in treatment protocols at other centers. Second, we do not have sufficient numbers for subgroup analysis by fracture type or postoperative immobilization protocols. At the same time, the authors feel this heterogeneity ensures the population studies is more representative of the average pediatric orthopaedist's clinical practice. Third, there was no follow-up beyond POD21. This decision reflected our hypothesis that most children would not need prolonged opioids, and in fact we were able to document that all children completed their course of opioids within this window. However, we acknowledge that patients may return to request pain medications in the long-term related to their lower extremity fracture, and this was not investigated in our protocol [17,37,38]. Fourth, we made a number of assumptions (see Methods) to account for missing survey data. While each of these assumptions was felt to be conservative, it nonetheless introduced a potential for error. Further studies with larger and more complete data sets will be required to corroborate our findings. Fifth, there are numerous confounding factors for surgery-related anxiety that are challenging to control for. For example, this study was unable to account for patients who went straight from the ED to surgery during the same admission (initial counseling by resident) versus those who were first seen in the outpatient clinic and had many days to process and prepare for surgery following (presumably) more thorough counseling directly with the attending surgeon. Sixth, many families are (justifiably) very stressed in a trauma situation while awaiting surgery, and may not have the time (or patience) to accurately and reliably fill out surveys. To account for this, we did encourage families to fill out surveys during the limited quiet time they experienced around surgery (e.g. while awaiting an end-of-day add-on after overnight ED admission, in the family waiting room during surgery, etc.) and even allowed them to fill it out up to 24 h after surgery. The authors also feel this best reflects the real-world scenarios many of our patients face on a daily basis. Additionally, a portion of our patients were “cold” trauma, thus allowing more perioperative time with which to complete surveys. Overall, trauma patients are a heterogenous and challenging population to study, but nonetheless important, as they make up a majority of many pediatric orthopaedist's practices. Seventh, this study relied on self-reported pill counts to quantify analgesic consumption, which may be inaccurate. However, this methodology has been previously validated in supracondylar humerus fractures, and we attempted to limit recall bias by texting families on a daily basis, only asking about usage for the past 24 h, and ultimately grouping responses into a “Yes” vs “No” category which is easier for families to remember. Finally, we acknowledge the limited number of patients surveyed here from a single institution. Although true normative data would certainly require larger numbers, we feel that the smaller numbers achieved through this pilot study are sufficient to provide preliminary counseling to patients and families, while also serving as a foundation from which to build larger (and potentially multicenter) studies further investigating these psychosocial risk factors across bigger populations.

In conclusion, this study contributes normative data for postoperative pain, analgesic consumption, and functional recovery following operative treatment of lower extremity fractures in children that can be used to guide prescribing practices and preoperative family education. Despite low pain self-efficacy and significant mobility limitations throughout the early postoperative period, most children did very well after surgery, speaking to their overall resilience. Nonetheless, baseline pain-related anxiety is a significant concern for many families, suggesting further work is necessary to determine how psychosocial variables might affect postoperative recovery and whether targeted interventions would be valuable. While reducing opioid prescriptions is of paramount importance across all of orthopaedics, it is equally important to identify and develop alternative evidence-based analgesic strategies so that we can safely balance opioid stewardship with quality patient care.

Additional links

• •JPOSNA®: Coping Skills in Children: An Introduction to the Biopsychosocial Model of Pain Control as a Tool to Improve Postoperative Outcomes
• •JPOSNA®: Variation in Prescribing Patterns and Use of Opioids in Pediatric Orthopaedic Surgery Patients

Consent for publication

Complete written informed consent was obtained from the patient, their guardian or legal representative for the publication of this study and accompanying images.

Author contributions

Alex L. Gornitzky: Writing – review & editing, Writing – original draft, Supervision, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Zoe E. Belardo: Writing – review & editing, Writing – original draft, Project administration, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Bridget K. Ellsworth: Writing – review & editing, Writing – original draft, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Sulagna Sarkar: Writing – review & editing, Writing – original draft, Formal analysis, Data curation, Conceptualization. Joseph L. Yellin: Writing – review & editing, Writing – original draft, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Arianna Trionfo: Writing – review & editing, Conceptualization. Apurva S. Shah: Writing – review & editing, Writing – original draft, Methodology, Formal analysis, Conceptualization.

Disclosures

None.

Funding

None.

Declaration of competing interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.