Disparities in Pediatric Orthopedic Surgery Care During the COVID-19 Pandemic Pre-vaccine and Post-Vaccine Availability

Background:

The COVID-19 pandemic has led to significant disruptions in medical care, resulting in an estimated 40% of US adults avoiding care. However, the return to baseline health care utilization following COVID-19 restrictions within the pediatric orthopedic population remains unexplored. We sought to analyze the visit volume and demographics of pediatric orthopedic patients at 3 timepoints: prepandemic (2019), pandemic (2020), and pandemic post-vaccine availability (2021), to determine the impact of COVID-19 restrictions on our single-center, multisite institution.

Methods:

We performed a retrospective cohort study of 6318 patients seeking treatment at our institution from May through August in 2019, 2020, and 2021. Patient age, sex, address, encounter date, and ICD-10 codes were obtained. Diagnoses were classified into fractures and dislocations, non–fracture-related trauma, sports, elective, and other categories. Geospatial analysis comparing incidence and geospatial distribution of diagnoses across the time periods was performed and compared with the Centers for Disease Control (CDC) social vulnerability index (SVI).

Results:

The total number of pediatric orthopedic visits decreased by 22.2% during the pandemic (P<0.001) and remained 11.6% lower post-vaccine availability compared with prepandemic numbers (P<0.001). There was no significant difference in age (P=0.097) or sex (P=0.248) of the patients across all 3 timepoints; however, patients seen during the pandemic were more often White race (67.7% vs. 59.3%, P<0.001). Post-vaccine availability, trauma visits increased by 18.2% (P<0.001) and total fractures remained 13.4% lower than prepandemic volume (P<0.001). Sports volume decreased during the pandemic but returned to prepandemic volume in the post-vaccine availability period (P=0.298). Elective visits did not recover to prepandemic volume and remained 13.0% lower compared with baseline (P<0.001). Geospatial analysis of patient distribution illustrated neighborhood trends in access to care during the COVID-19 pandemic, with fewer patients from high SVI and low socioeconomic status neighborhoods seeking fracture care during the pandemic than prepandemic. Post-vaccine availability, fracture population distribution resembled prepandemic levels, suggesting a return to baseline health care utilization.

Conclusion:

Pediatric orthopedic surgery visit volume broadly decreased during the COVID-19 pandemic and did not return to prepandemic levels. All categories increased in the post-vaccine availability time point except elective visits. Geospatial analysis revealed that neighborhoods with a high social vulnerability index (SVI) were associated with decreased fracture visits during the pandemic, whereas low SVI neighborhoods did not experience as much of a decline. Future research is needed to study these neighborhood trends and more completely characterize factors preventing equitable access to care in the pediatric orthopedic population.

Level of Evidence:

Retrospective Study, Level III.

The COVID-19 pandemic has led to significant disruptions in medical care.^1–7 As of October 2022, over 1 million Americans have died from SARS-CoV-2, and an estimated 40% of US adults reported avoiding medical care during the pandemic.^8,9 Importantly, delays in care have been associated with education level, household income, and zip code.^10–12 Therefore, COVID-19-related medical care delay or avoidance may further increase morbidity and mortality for treatable and preventable health outcomes, exacerbating existing health inequalities in the United States.

The pediatric orthopedic population is particularly vulnerable to socioeconomic factors impacting preventative, routine, and follow-up care.^13,14 Compliance in rehabilitation is essential for common orthopedic conditions such as sports-related trauma and fractures.^14,15 Variables such as single-parent status, household income, and distance to hospital have been associated with decreased rehabilitation compliance in pediatric patients, suggesting a limitation in effective health care delivery for these populations.¹⁶ Geospatial analysis may be employed to identify high-priority populations experiencing adverse health outcomes¹⁷; however, this analysis has not yet been applied to address existing pediatric orthopedic health disparities, nor has it been used to identify asymmetric impacts on care delivery due to the COVID-19 pandemic.

Preliminary studies found that COVID-19 restrictions significantly decreased the incidence of common pediatric fractures associated with sporting and outdoor activities, consistent with an increase in at-home causes of injury.^18,19 However, the impact of COVID-19 on patients seeking elective care is unknown. In addition, the return to baseline following COVID-19 restrictions (post-vaccine availability) on the pediatric orthopedic population remains unexplored. Broadly, no study to date has comprehensively evaluated the health disparities associated with COVID-19-mediated reduction in pediatric orthopedic surgery volume.

Considering the existing prepandemic disparities in access to orthopedic care, understanding associations between socioeconomic status (SES), patient residence, and care during the pandemic may allow for interventions that may be designed at several levels to mitigate existing disparities. Here, we analyze the visit volume, demographics, and geospatial location of pediatric orthopedic patients in 3 timepoints, 2019, 2020, and 2021 to determine the impact of COVID-19 restrictions on our single-center, multisite institution. We hypothesized that the number of patients presenting to our institution would decrease during the pandemic and remain below prepandemic rates in the post-vaccine availability timepoint.

METHODS

We performed an institutional review board-approved retrospective cohort study investigating visit volume, demographics, and geospatial location of pediatric orthopedic patients across 3 timepoints to determine the impact of COVID-19 restrictions on access to care. All patients seen in the Division of Pediatric Orthopedics (outpatient and emergency room) at a single-center, multisite pediatric level 1 trauma center from May to August in 2019, 2020, and 2021 were screened, comprising “prepandemic,” “pandemic,” and “post-vaccine availability,” respectively. Although both 2020 and 2021 timepoints are still considered “pandemic” according to the CDC, 2021 was defined by the vaccine availability in our study to reflect population-level shifts following widespread vaccination. The 3-month window from May to August was selected to avoid heterogeneous data during the early months (March, April) of COVID lockdowns in the US, and to capture a representative sample of spring and summer orthopedics visits. In addition, the FDA approved vaccination for children 12 years of age and older in May 2021. Patient age, sex, address, zip code, encounter date, and ICD-10 codes were obtained with assistance from eResearch and stored in a secure database. Age categories were created based on the Centers for Disease Control child development groupings.²⁰

Diagnoses were clinically classified into fractures and dislocations; nonfracture acute trauma; sports; and elective; all remaining diagnoses were classified as “other.” We categorized nontraumatic, nonurgent, and nonfracture orthopedic visits into elective visits. Within elective visits, we distinguished between spine, congenital, upper extremity, lower extremity, and other visits. We classified sports visits as either “overuse” related or sprain/strain categories. ICD-10 codes and ICD descriptions were manually consolidated into clinically meaningful categories (ie, wrist fracture) without respect for laterality.

Categorical variables were compared utilizing χ² or Fisher exact tests, whereas continuous variables were compared using a Z-test for 2 population proportions to test the null hypothesis that the 2 populations were not different during each specified year (2019 vs. 2020, 2020 vs. 2021, and 2021 vs. 2019). All data analysis was performed in Microsoft Excel and R (R 4.2.1 binary macOS).

Geospatial Analysis

Patient address, city, state, and postal code were converted to geographical coordinates and mapped in a Geographic Information System (GIS), ArcGIS (Esri, CA) previously described.²¹ Duplicate patient IDs were also filtered out for each year (2019, 2020, and 2021). Using a point-in-polygon approach, patient addresses were associated with the social vulnerability index (SVI). The SVI is a combination of different social, economic, and environmental measures used by the CDC to geographically represent variations in risk and opportunity and is reported as a value between 0 (low) and 1 (high) vulnerability for each census tract.²²

A series of attribute manipulations and cartographic visualizations were then created within ArcGIS to show geographic variation between each time period. These data were visualized for exploratory purposes using a Kernel Density Estimation, which creates a heat surface map of high intensities not limited by census tract boundaries. This commonly used method in spatial epidemiology was utilized for COVID-19 surveillance during the pandemic.

RESULTS

In total, 6318 patients were included in this study. The absolute number of pediatric orthopedic visits decreased by 22.2% during the pandemic (2943 vs. 2289, P<0.001, Fig. 1, Table 1). There was a 13.7% increase in overall visits post-vaccine availability (2603 vs. 2289, P<0.001, Fig. 1A); however, post-vaccine volume remained 11.6% lower than prepandemic baseline (2943 vs. 2603, P<0.001). The mean presenting age for all patients increased slightly in the post-vaccine availability timepoint; however, this change was not statistically significant (10.6±5.1 vs. 11.0±5.9, P=0.293). There were no statistically significant differences in sex (P=0.097), although a greater proportion of patients were White during the pandemic compared with prepandemic (67.7% vs. 59.3%, P<0.001, Table 1).

FIGURE 1.

Yearly change in pediatric orthopaedic fracture, elective, trauma, and sports visits (A) and average distribution of visits from 2019 to 2021 (B).

TABLE 1.

Pediatric Orthopedic Surgery Patient Demographics and Case Volume During the COVID-19 Pandemic Pre-vaccine and Post-vaccine Availability

Variables	Prepandemic (2019)	Pandemic (2020)	Post-vaccine availability (2021)	P (2019 vs. 2020)	P (2020 vs. 2021)	P (2019 vs. 2021)
Total	2943	2289	2603	<0.001	<0.001	<0.001
Sex	—	—	—	0.906	0.248	0.267
Male	1582 (53.7)	1227 (53.6)	1439 (55.2)	—	—	—
Female	1361 (46.2)	1062 (46.4)	1164 (44.7)	—	—	—
Mean age	10.6±5.1	10.6±5.2	11.0±5.9	0.168	0.097	0.293
≤5	578	487	489	—	—	—
6–11	1086	793	931	—	—	—
≥12	1279	1009	1183	—	—	—
Race	—	—	—	<0.001	<0.001	0.054
White	1748 (59.3)	1550 (67.7)	1626 (62.4)	—	—	—
Black	915 (31.0)	531 (23.1)	744 (28.6)	—	—	—
Hispanic/Latino	118 (4.0)	89 (3.8)	83 (3.2)	—	—	—
Multiracial/Other	111 (3.7)	82 (3.6)	112 (4.3)	—	—	—
Asian	25 (0.8)	25 (1.0)	24 (0.9)	—	—	—
Unknown/declined	26 (0.8)	12 (0.5)	14 (0.5)	—	—	—
Total fractures	1434	1050	1241	<0.001	<0.001	<0.001
Upper Extremity Fractures	949 (66.1)	672 (64.0)	822 (66.2)	<0.001	<0.001	<0.001
Lower Extremity Fractures	308 (21.4)	241 (22.9)	282 (22.7)	<0.001	0.011	0.131
Dislocations/Subluxations	95	63	63	<0.001	1	<0.001
Total trauma (nonfracture)	181	140	202	0.002	<0.001	0.118
Violence	91 (50.0)	84 (60.0)	87 (43.0)	0.453	0.749	0.682
Motor vehicle	84 (46.4)	48 (34.2)	100 (49.5)	<0.001	<0.001	0.095
Total sports	272	182	255	<0.001	<0.001	0.298
Sprain or strain	158 (58.0)	80 (43.9)	109 (42.7)	<0.001	0.003	<0.001
Total elective	837	754	728	0.004	0.343	<0.001
Spine	364 (43.4)	366 (48.5)	316 (43.4)	0.92	0.009	0.012

Fractures

The total number of pediatric fractures and dislocations decreased by 26.7% during the pandemic compared with prepandemic (1050 vs. 1434, P<0.001, Supplementary Table 1, Supplemental Digital Content 1, http://links.lww.com/BPO/A640). Total fracture visits increased by 18.1% in the post-vaccine availability time point (1241 vs. 1050, P<0.001); however, this was still 13.4% lower than prepandemic (P<0.001). The most common fractures seen are summarized in Table 1 and Supplementary Table 1, Supplemental Digital Content 1, http://links.lww.com/BPO/A640.

Trauma Excluding Fractures and Dislocations

The absolute number of pediatric orthopedic visits for nonfracture trauma decreased by 22.6% during the pandemic (140 vs. 181, P<0.002, Supplementary Table 2, Supplemental Digital Content 2, http://links.lww.com/BPO/A641). There was a corresponding 44.6% increase in visits for trauma post-vaccine (202 vs. 140, P=<0.001). There was an 11.6% increase in post-vaccine trauma visits compared with prepandemic; however, this association was not statistically significant (P=0.118). Motor vehicle injury visits decreased by 42.9% during the pandemic (P<0.001) and rose by 108% post-vaccine availability (P<0.001), representing a 19.0% increase compared with prepandemic levels. Pedestrian accidents increased by 272% post-vaccine availability compared with the pandemic time point (41 vs. 11, P<0.001, Supplementary Table 2, Supplemental Digital Content 2, http://links.lww.com/BPO/A641). All remaining nonfracture trauma diagnoses are reported in Supplementary Table 4, Supplemental Digital Content 4, http://links.lww.com/BPO/A643.

Sports

The absolute number of pediatric orthopedic visits for sports-related injuries decreased by 33.0% during the pandemic (182 vs. 272, P<0.001, Supplementary Table 3, Supplemental Digital Content 3, http://links.lww.com/BPO/A642). There was a corresponding 40.1% increase in sports visits post-vaccine availability (255 vs. 182, P<0.001), indicating a return to baseline levels comparable with prepandemic time point (P=0.298). Among the overuse visits, we observed a 44.4% increase during the pandemic (n=52 vs. n=36, P=0.016). Pediatric sprain/strain visits saw a 49.3% decrease during the pandemic (80 vs. 158, P<0.001) and remained 31.0% below prepandemic levels at the post-vaccine availability time point (109 vs. 158, P<0.001).

Elective

The absolute number of pediatric orthopedic visits for elective visits decreased by 9.9% during the pandemic (754 vs. 837, P=0.004, Table 2). We observed a continued decline in elective visits during the post-vaccine availability time point, corresponding to a 13.0% decrease in 2021 compared with 2019 (728 vs. 837, P<0.001, Table 2).

TABLE 2.

Elective Case Volume During the COVID-19 Pandemic Pre-vaccine and Post-vaccine Availability

Variables	Prepandemic (2019)	Pandemic (2020)	Post-vaccine availability (2021)	P (2019 vs. 2020)	P (2020 vs. 2021)	P (2019 vs. 2021)
Total	837	754	728	0.004	0.322	<0.001
Spine	364	366	316	0.92	0.009	0.012
Adolescent idiopathic scoliosis	166	206	153	0.003	<0.001	0.303
Juvenile idiopathic scoliosis	87	84	82	0.749	0.826	0.589
Infantile idiopathic scoliosis	12	8	9	0.208	0.728	0.352
Neuromuscular scoliosis	57	38	42	0.006	0.529	0.033
Congenital deformity of spine	29	17	17	0.009	0.596	0.002
Spondylolisthesis/Spondylolysis/Spondylosis	13	13	13	1	1	1
Congenital	69	69	64	0.719	0.472	0.711
Developmental dysplasia of the hip	14	9	6	0.142	0.271	0.011
Congenital metatarsus adductus	7	9	5	0.478	0.131	0.412
Congenital pes cavus	7	0	13	<0.001	<0.001	0.057
Congenital pes planus	6	6	2	1	0.046	0.046
Congenital talipes calcaneovalgus	0	3	5	0.014	0.317	0.002
Congenital talipes equinovarus	35	42	33	0.258	0.142	0.728
Upper extremity	5	4	3	0.638	0.596	0.317
Mallet finger	2	2	1	1	0.412	0.412
Trigger finger	3	2	2	0.529	1	0.529
Lower extremity	289	224	258	<0.001	0.029	0.061
Acquired clubfoot	19	15	5	0.332	0.002	<0.001
Acquired flat foot (pes planus)	49	39	37	0.131	0.749	0.067
Hallux valgus	21	15	11	0.159	0.267	0.012
Trochanteric bursitis	4	6	0	0.373	<0.001	0.005
Short Achilles tendon	115	88	139	0.007	<0.001	0.033
Baker cyst	5	2	3	0.11	0.529	0.317
Unequal limb length	76	59	63	0.038	0.61	0.119
Other
Accessory digit	8	5	6	0.238	0.667	0.447
Bone cyst	4	3	7	0.596	0.073	0.201
Benign neoplasm of bone	19	28	27	0.063	0.849	0.095
Cervicalgia	14	4	7	0.001	0.201	0.031
Juvenile osteochondrosis	65	51	40	0.066	0.103	<0.001

Patients being seen for spinal conditions were the largest contributor to the elective caseload at our center, comprising 43.4%, 48.4%, and 43.3% of elective visits in 2019, 2020, and 2021, respectively. Among elective spine surgery visits, there was no significant decline during the pandemic (366 vs. 364, P=0.92). Spine visits declined by 13.2% in 2021 compared with 2019 (316 vs. 364, P=0.009). The most common spine diagnoses seen during the pandemic were adolescent idiopathic scoliosis (AIS, 56.2%) and juvenile idiopathic scoliosis (JIS, 22.9%). AIS visits increased by 24.1% during the pandemic (206 vs. 166, P=0.003, Table 2).

Geospatial Analysis

To evaluate the social determinants impacting pediatric patients’ access to health during the COVID-19 pandemic, we performed a geospatial analysis using kernel density estimate heat maps (Fig. 2). As shown in Figure 2A, the greatest density of pediatric patients in 2019 was centered around our hospital’s main campus, broadly covering the east side of Cleveland and adjacent suburbs. Two other hotspots in 2019 included Euclid and Parma. Figure 2C demonstrates the heat map density of pediatric patients in 2020 during the pandemic. The hotspot is primarily centered on the east side suburbs of Cleveland and includes a reduced density of downtown and eastside Cleveland residences. In 2021 (Fig. 2D), the distribution resembles 2019 more closely, with the heat map encompassing these neighborhoods again. We then referenced the CDC SVI to investigate whether there was any association between reduced care in 2020 and increased social vulnerability (Fig. 2B). Visually, the neighborhoods that experienced decreased levels of pediatric orthopedic care during the pandemic were directly associated with heightened social vulnerability.

FIGURE 2.

Geospatial disparities in pediatric orthopaedic fracture patients across Cleveland. Kernel density estimate heat map of fracture visits in 2019 (A), 2020 (C), and 2021 (D) reveal trends in neighborhood access to care during the pandemic. CDC Social Vulnerability Index based on socioeconomic status, household composition, minority status and language, and housing and transportation illustrates the association between high social vulnerability neighborhoods and decreased utilization of care during the pandemic (B).

DISCUSSION

The present study reports the impact of the COVID-19 pandemic on the return to baseline levels of care in the pediatric orthopedic population post-vaccine availability. Our results demonstrated that the total number of pediatric orthopedic visits remained 11.6% lower in 2021 compared with 2019, suggesting that our single-site, multicenter institution did not yet return to baseline levels. This study has significant clinical implication for pediatric orthopedic surgery centers in academic medical settings seeking to understand the demographics and shift in patient needs since the pandemic in 2020.

During the onset of COVID-19, in-person school and workplace activities were cancelled, resulting in significantly increased at-home time for children aged 0 to 18. In this study, we report that pediatric fracture and dislocation visits declined by 26.7% during the pandemic compared with prepandemic. Our results are consistent with a recent study investigating pediatric fracture incidence in Philadelphia.¹⁸ We suspect that this significant decline is due to multiple factors. One leading hypothesis is simply that less fractures occurred. With decreased activity (eg, less organized school sports), children had decreased opportunity to fracture bones. As expected, this coincided with an increase in the proportion of fractures occurring at home.¹⁸ Another hypothesis is that patients and families were less likely to seek care during the pandemic.²³ Fractures may have occurred at similar rates but went untreated during this time period. If true, we may expect to see long-term musculoskeletal consequences secondary to untreated fractures in these patients. In this study, we determined that after the vaccine was distributed, fracture care visits increased by 18.1% but still remained 13.5% below prepandemic levels. We suspect the maintained decrease in 2021 compared with 2019 is due to 3 factors: (1) different behavior due to the ongoing pandemic, (2) decreased health care utilization, and (3) caution when utilizing hospital and physician services due to the ongoing pandemic.

There are many factors at play when considering patient utilization of health care resources. Studies suggest that cost, time, and distance to practice all significantly influence the likelihood of seeing a physician.^12,14 In pediatric orthopedics, these findings have been reproduced, with evidence suggesting that those with public health insurance may delay care more than those with private health insurance, possibly due to cost concerns.¹⁴ In our patient population, despite being unable to directly assess insurance status or SES, we were able to infer patient SES using the CDC (SVI and map these results using geospatial analysis.²⁴

In this study, we observed that increased SVI neighborhoods were associated with a pronounced decrease in visits to our system during the pandemic. The only significant demographic change during the pandemic was a greater proportion of patients of the White race seen during the pandemic. These results highlight the complex nature of social determinants of health on access to care. Greater efforts are needed to target patients from high SVI neighborhoods; for example, mobile fracture clinics and clinical follow-up posttreatment to ensure disparities in care are minimized.

For nonfracture trauma visits, visit volume stabilized in 2021 compared with 2019. Despite expecting that violent injuries would increase during the pandemic due to potential nonaccidental trauma being in unsafe home environments, we observed a 7.7% decrease. We hypothesize that these results are due to decreased health care utilization during the pandemic. In support of this hypothesis, studies have found rates of at-home violence during the pandemic were 8 times higher than expected, suggesting our results may be explained by patient decision rather than trauma incidence.²⁵ Our study also observed a 44% increase in overuse visits during the pandemic. One possible hypothesis is that patients injured themselves with classic “overuse” injuries such as sprain/strain due to de-conditioning during virtual school.

For elective visits, one hypothesis could be that patients would seek care at higher rates during the pandemic due to increased work and school flexibility. We found instead that elective visits declined during the pandemic and continued to decrease post-vaccine timepoint. The exact cause of this decline is uncertain but is likely related to the fear of contracting COVID. Of note, adolescent idiopathic scoliosis visits increased during the pandemic, suggesting patients may have sought care at higher rates in 2020 due to being homebound. Future studies are needed to investigate the relationship between elective and trauma health care utilization and its association with household SES, insurance status, and SVI.

This study has several limitations. First, as a retrospective cohort study, we are limited in generalizing to the general population. Second, utilizing the electronic medical record has inherent drawbacks, which are well-described in the literature.²⁶ Third, there may be residual overlap between categories. However, our categorization methodology using ICD-10 codes minimized this overlap, and our large sample size of over 6000 patients should reduce variance from individual clinician decision-making in aggregate. In addition, our post-vaccine availability reflects the period immediately following vaccine distribution and is not a reflection of a true “postpandemic” timepoint. Finally, our methodology did not distinguish between surgical versus outpatient visits; therefore we are limited in making statements about operative usage across 3 timepoints.

In conclusion, our results demonstrated that the total number of pediatric orthopedic visits remained 11.6% lower in 2021 compared with 2019, suggesting that our single-site, multicenter institution did not yet return to baseline levels. Notably, pediatric visits for fracture care increased substantially post-vaccine availability but remained below prepandemic levels. In contrast, we observed a steady decline in pediatric visits for elective care. The pandemic significantly affected patients’ ability to come to the hospital due to lack of resources, transportation, and fear of contracting COVID, among other factors. Our study is the first to report an association between SVI and population-level changes in the pediatric orthopedic population, making this an important contribution to the literature in our field. Future studies are needed to fully characterize disparities in access and utilization of health care in the pediatric orthopedic population, especially comparing elective and traumatic causes.

ACKNOWLEDGMENTS

The authors thank Abdus Sattar and Chenya Zhao for their statistical assistance.

Supplementary Material

bpo-43-529-s001.docx ^{(18.5KB, docx)}

bpo-43-529-s002.docx ^{(16.2KB, docx)}

bpo-43-529-s003.docx ^{(17.3KB, docx)}

bpo-43-529-s004.docx ^{(15KB, docx)}