Association of Cerebral Palsy With Unanticipated Admission Following Pediatric Ambulatory Surgery
Funding: The authors received no specific funding for this work.
ABSTRACT
Background
Cerebral palsy, a neurologic disorder caused by damage to the developing brain, is a leading cause of childhood disability. Due to musculoskeletal, movement, and secondary impairments, children with cerebral palsy often require surgical care. With the growing cost of surgical care, many children with cerebral palsy are scheduled for surgery in an ambulatory setting. Whether cerebral palsy increases the risk of unanticipated admission (a critical quality indicator of care) following ambulatory surgery has not been characterized. Our objective was to determine the association of cerebral palsy with unanticipated admission following pediatric ambulatory surgery.
Methods
We used the Pediatric Health Information System (PHIS) database to evaluate a retrospective cohort of children (< 18 years) who underwent scheduled ambulatory operations between January 1, 2010 and December 31, 2022. The primary outcome was unanticipated admission. Using log-binomial regression models, we estimated the relative risk and 95% confidence intervals for unanticipated admission, comparing patients with and without cerebral palsy. To account for confounding variables, we performed a 1:1 propensity score matching without replacement.
Results
A total of 1 954 108 children underwent ambulatory surgeries during the study period. Of these, 4.1% required unanticipated admission. The overall incidence of unanticipated admission was significantly higher among children with cerebral palsy than in those without (9.8% vs. 4.0%; p < 0.001). This association remained significant after multivariable adjustment (relative risk: 1.73; 95% CI: 1.59–1.87, p < 0.001).
Conclusion
Although cerebral palsy is not a contraindication for ambulatory surgery in children, it is significantly associated with the risk of unanticipated hospital admissions. This underscores the need for careful preoperative clinical site of care selection in this vulnerable patient population.
Level of Evidence
Level II.
Abbreviations
-
- APR-DRG
-
- all patient refined diagnosis related groups
-
- ASA
-
- American Society of Anesthesiologists
-
- CI
-
- confidence interval
-
- CP
-
- cerebral palsy
-
- ICD
-
- International Classification of Diseases
-
- PHIS
-
- Pediatric Health Information System
-
- RR
-
- relative risk
1 Introduction
Cerebral palsy (CP) is a group of neurological disorders caused by insult to the developing brain, which leads to permanent motor dysfunction and impaired ability to maintain balance and posture [1]. It is the leading cause of chronic childhood disability with an estimated incidence of 2 per 1000 live births in the United States [2]. CP is frequently associated with many other conditions, such as intellectual disabilities, seizure disorders, scoliosis, chronic pain, sensory processing challenges, malnutrition, gastroesophageal reflux, and chronic pulmonary disease [3]. Furthermore, children with CP often require frequent, sometimes recurrent surgical procedures including spinal fusion, pelvic or femoral osteotomy, tendon lengthening or release, and botulinum toxin injections to reduce spasticity. Other interventions include insertion of a baclofen pump, strabismus repair, placement of a vagal nerve stimulator, gastrostomy tube insertion, endoscopy, dental rehabilitation, and deep brain stimulation [4]. Many of these procedures may be performed on an ambulatory basis, prompting the need to evaluate the safety and appropriateness of scheduling these medically complex children for ambulatory perioperative care.
Ambulatory surgery represents over two-thirds of the surgeries performed in the United States each year [5], amounting to roughly 12 million procedures in 2019 [6]. Ambulatory surgery offers significant benefits to patients and the healthcare system compared to inpatient surgery [5]. Ambulatory surgeries decrease the risk of patient exposure to nosocomial pathogens, reduce iatrogenic injuries, improve patient satisfaction, and decrease healthcare resources [7]. Furthermore, ambulatory surgeries incur significantly lower cost of care compared to inpatient operations, with estimates indicating reductions ranging from 17% to 43% of hospital costs [8, 9].
Despite the benefits of ambulatory surgery, poor patient selection or complications during surgery and anesthesia can lead to unscheduled admissions. These admissions disrupt patients and their families, impact hospital workflow, strain resources, and result in patient dissatisfaction [10]. Unanticipated admission occurs following approximately 2% of pediatric ambulatory surgeries [11].
Given the rising frequency of pediatric ambulatory surgery and the prevalence of cerebral palsy (CP) in the general population, this study aimed to determine the association of CP with unanticipated admissions following ambulatory surgery. Understanding this association may guide decisions about the suitability of ambulatory procedures for this patient population. We hypothesized that the risk of unanticipated admission following ambulatory surgery is higher in children with CP compared to those without.
2 Materials and Methods
2.1 Study Design and Data Source
We performed a retrospective cohort study utilizing data from 2010 to 2022 from the Pediatric Health Information System (PHIS) database. The PHIS collects de-identified discharge-level data from 49 freestanding, tertiary children's hospitals affiliated with the Children's Hospital Association [12]. PHIS reports individual characteristics, including sociodemographic information, International Classification of Disease codes, and all patient refined diagnosis related groups (APR-DRG) codes. Additionally, hospitals report date-stamped billing data using Clinical Transaction Classification codes (IBM Watson, Armonk, NY). Data reporting is subject to consistency reviews, hospital-specific data quality audits, coding consensus meetings, and data quality reports. The institutional review board at our institution approved the study protocol and waived the requirement for written informed consent. We adhered to Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines for cohort studies.
2.2 Study Population
Our analysis included children < 18 years of age who underwent surgery scheduled as ambulatory status. Patients with CP were identified using ICD 9 codes 333.71, 343.0–4, and 343.8–9 and ICD 10 codes G80.0-G80.4, G80.8-G80.9, as previously described [13, 14].
2.3 Study Outcomes
The primary outcome was unanticipated admission, defined as postoperative stay ≥ 1 day. This definition is consistent with previous studies among pediatric and adult ambulatory surgical populations [15, 16].
2.4 Statistical Analyses
Baseline differences between children with and without CP who required unanticipated admission were described by the frequency (percentage). We estimated the relative risk (RR) and 95% confidence intervals (CI) for unanticipated admission comparing children with and without CP using log-binomial regression models. To account for confounding, we performed a 1:1 propensity score matching without replacement. We used generalized estimating equations (GEE) to account for the matched sample structure, with each matched pair considered a cluster. GEE is a statistical method used in analyses that have correlated data (e.g., clustered data) [17, 18]. We estimated the propensity score of having CP using multivariable logistical regression that included the following variables: age (≤ 5, 6–12, or ≥ 13 years of age), sex (male vs. female), insurance status (commercial, Medicaid, other), presence of key chronic complex conditions (cardiovascular, gastrointestinal, hematologic or immunologic, malignancy, metabolic, renal and urologic, respiratory, ADHD, autism, congenital malformations of the nervous system, epilepsy, nervous and musculoskeletal system), and procedural group (craniofacial/plastics, gastroenterology, nephrology/urology, ophthalmology, orthopedics, otolaryngology, or vascular). We defined complex chronic conditions according to Feudtner et al. as “any medical condition that can be reasonably expected to last at least 12 months (unless death intervenes) and to involve either several different organ systems or 1 organ system severely enough to require specialty pediatric care and probably some period of hospitalization in a tertiary care center.” [19] To measure the imbalance in baseline characteristics between children with and without CP, we calculated the absolute standardized difference in proportion [20]. A standardized difference ≥ 0.10 indicated a meaningful difference between cohorts. A variance inflation factor < 10 was considered to exclude significant multicollinearity. We evaluated the extent of missing values for all variables. (Table S1) Given that missingness was deemed minimal (< 5%), we performed complete case analysis. Also, among patients with CP, we sought to identify characteristics that were independently associated with unanticipated admission. We used logistic regression to build a prognostic model for unanticipated admission among children with CP. Using a purposeful framework [21], a variable was included into the multivariable model if its crude association with unanticipated admission was significant at the alpha level of 0.20 [22]. A p value of ≤ 0.05 was considered statistically significant. All statistical analyses were performed using STATA version 16 (StataCorp, College Station, TX: Stata Press).
3 Results
3.1 Study Population Characteristics
There were 1 954 108 children under the age of 18 who underwent surgeries designated as ambulatory status at a PHIS-reporting institution between January 1, 2010 and December 31, 2022. Of these, 15 676 (0.8%) had a diagnosis of CP. The majority of patients with CP were in the 6–12 year age group (44.3% of the CP cohort), compared to patients without CP, the majority of whom were in the 5 year or less age group (57.9%). Patients with CP were more likely to be insured by Medicaid (51.5% vs. 40.6%) and from ZIP codes with lower median income. They more commonly had cardiovascular (2.5% vs. 1.1%), gastrointestinal (19.2% vs. 1.2%), metabolic (1.8% vs. 0.5%), and respiratory (6.0% vs. 1.3%) chronic complex conditions than children without CP. Additionally, children with CP were more likely to have comorbid ADHD, autism, congenital malformations of the nervous system, epilepsy, and nervous and musculoskeletal system disorders. Children with CP more commonly underwent orthopedic procedures (37.1% vs. 15.2%) and ophthalmology procedures (24.9% vs. 11.8%) than children without CP. After propensity score matching, the cohorts of children with and without CP were comparable with respect to baseline characteristics (absolute standardized difference in proportion < 0.10) (Table 1).
Characteristics | Before propensity score matching | After propensity score matching | ||||
---|---|---|---|---|---|---|
Cerebral palsy | Std diff | Cerebral palsy | Std diff | |||
No | Yes | No | Yes | |||
Study population | 1 938 432 (99.2) | 15 676 (0.80) | 15 465 (50.0) | 15 465 (50.0) | ||
Male sex | 1 169 879 (60.4) | 9346 (59.6) | 0.02 | 9032 (59.1) | 9091 (59.5) | 0.01 |
Age | ||||||
≥ 13 | 274 935 (14.2) | 3255 (20.8) | 0.17 | 3071 (20.1) | 3150 (20.6) | 0.01 |
6–12 | 540 607 (27.9) | 6940 (44.3) | 0.35 | 6506 (42.6) | 6755 (44.2) | 0.03 |
≤ 5 | 1 122 890 (57.9) | 5481 (35) | 0.47 | 5701 (37.3) | 5373 (35.2) | 0.04 |
Insurance status | ||||||
Commercial | 985 119 (51.9) | 6511 (41.9) | 0.20 | 6482 (42.4) | 6426 (42.1) | 0.01 |
Medicaid | 771 122 (40.6) | 7993 (51.5) | 0.22 | 7861 (51.5) | 7848 (51.4) | 0.00 |
Other | 141 946 (7.5) | 1027 (6.6) | 0.03 | 935 (6.1) | 1004 (6.6) | 0.02 |
Chronic complex condition | ||||||
Cardiovascular | 20 836 (1.1) | 397 (2.5) | 0.11 | 470 (3.1) | 380 (2.5) | 0.04 |
Gastrointestinal | 23 697 (1.2) | 3008 (19.2) | 0.62 | 2747 (18) | 2743 (18) | 0.00 |
Hematologic or Immunologic | 13 196 (0.7) | 247 (1.6) | 0.08 | 248 (1.6) | 236 (1.5) | 0.01 |
Malignancy | 20 940 (1.1) | 79 (0.5) | 0.07 | 65 (0.4) | 77 (0.5) | 0.01 |
Metabolic | 9051 (0.5) | 287 (1.8) | 0.13 | 287 (1.9) | 276 (1.8) | 0.01 |
Renal and Urologic | 21 170 (1.1) | 345 (2.2) | 0.09 | 355 (2.3) | 325 (2.1) | 0.01 |
Respiratory | 24 503 (1.3) | 934 (6) | 0.25 | 983 (6.4) | 880 (5.8) | 0.03 |
ADHD | 35 854 (1.8) | 584 (3.7) | 0.11 | 629 (4.1) | 569 (3.7) | 0.02 |
Autism | 18 192 (0.9) | 732 (4.7) | 0.23 | 775 (5.1) | 706 (4.6) | 0.02 |
Congenital malformations of the nervous | 11 059 (0.6) | 1392 (8.9) | 0.40 | 1370 (9) | 1270 (8.3) | 0.02 |
Epilepsy | 10 569 (0.5) | 3316 (21.2) | 0.70 | 2954 (19.3) | 3039 (19.9) | 0.01 |
Nervous and musculoskeletal system | 7368 (0.4) | 682 (4.4) | 0.26 | 713 (4.7) | 645 (4.2) | 0.02 |
Procedural group | ||||||
Craniofacial/Plastic | 43 805 (2.4) | 188 (1.2) | 0.09 | 180 (1.2) | 187 (1.2) | 0.00 |
Gastroenterology | 34 250 (1.8) | 330 (2.1) | 0.02 | 313 (2) | 318 (2.1) | 0.00 |
Nephrology/Urology | 239 912 (12.9) | 1183 (7.5) | 0.18 | 1068 (7) | 1134 (7.4) | 0.02 |
Ophthalmology | 218 951 (11.8) | 3902 (24.9) | 0.34 | 3972 (26) | 3806 (24.9) | 0.02 |
Orthopedics | 282 770 (15.2) | 5815 (37.1) | 0.51 | 5493 (36) | 5654 (37) | 0.02 |
Otolaryngology | 1 032 389 (55.5) | 4210 (26.9) | 0.61 | 4200 (27.5) | 4132 (27) | 0.01 |
Vascular | 8410 (0.5) | 48 (0.3) | 0.02 | 52 (0.3) | 47 (0.3) | 0.01 |
3.2 Association of Cerebral Palsy With Unanticipated Admission After Ambulatory Surgery
Overall, 4.1% of the children undergoing ambulatory surgery during the study period required unanticipated admission ≥ 1 day. On univariable analysis, unanticipated admission rate was significantly higher in children with comorbid CP than in those without CP (9.8% vs. 4.0%; p < 0.001). After adjusting for age, sex, insurance status, procedural group, and comorbidities, CP was associated with nearly twice the risk of unanticipated admission—RR 1.73 (95% CI: 1.59–1.87, p < 0.001) (Table 2).
Cerebral palsy | Unanticipated admission n/N (%) | Unadjusted | Propensity score adjusted | ||
---|---|---|---|---|---|
RR (95% CI) | p | RR (95% CI) | p | ||
No | 77 952/1 938 432 (4.0) | Reference | Reference | ||
Yes | 1535/15 676 (9.8) | 2.43 (2.32–2.55) | < 0.001 | 1.73 (1.59–1.87) | < 0.001 |
- Note: Adjusted analyses were controlled for male sex, race, age, insurance status, procedural group, and the presence of key preoperative comorbidities: cardiovascular, gastrointestinal, hematologic or immunologic, malignancy, metabolic, renal and urologic, respiratory, ADHD, autism, congenital malformations of the nervous system, epilepsy, and nervous and musculoskeletal.
- Abbreviations: CI, confidence interval; RR, relative risk.
3.3 Factors Associated With Unanticipated Admission After Ambulatory Surgery in Children With Cerebral Palsy
Among children with CP, younger age (< 12 years), insurance status, epilepsy, and procedural group were independently associated with unanticipated admission. Of the procedures, the greatest risk of unanticipated admission was following vascular and orthopedic procedures; ophthalmology procedures were least likely to result in unanticipated admission. (Table 3).
Characteristics | Relative risk (95% CI) | p |
---|---|---|
Age | ||
≥ 13 | Reference | |
6–12 | 1.13 (0.99–1.29) | 0.06 |
≤ 5 | 1.32 (1.14–1.54) | < 0.01 |
Insurance status | ||
Commercial | Reference | |
Medicaid | 0.63 (0.57–0.70) | < 0.01 |
Other | 0.42 (0.31–0.55) | < 0.01 |
Epilepsy | ||
No | Reference | |
Yes | 1.16 (1.03–1.32) | 0.02 |
Procedural group | ||
Ophthalmology | Reference | |
Craniofacial/Plastic | 7.05 (4.01–12.38) | < 0.01 |
Gastroenterology | 7.32 (4.69–11.45) | < 0.01 |
Nephrology/Urology | 2.70 (1.78–4.09) | < 0.01 |
Orthopedics | 14.02 (10.56–18.61) | < 0.01 |
Otolaryngology | 6.13 (4.57–8.23) | < 0.01 |
Vascular | 16.92 (8.57–33.38) | < 0.01 |
4 Discussion
We leveraged a large nationally representative pediatric sample from the past 12 years to evaluate the association of comorbid CP with unanticipated admission following ambulatory surgery. We found that, although the overall prevalence of CP in children who underwent ambulatory surgery was low, CP had a significant effect on the primary outcome of unanticipated admission. Specifically, patients with CP had nearly twice the risk of unanticipated admission than patients without CP. Due to musculoskeletal and other systemic disorders, children with CP often require surgery. With the growing cost of healthcare and a relentless push for ambulatory care, many children with CP are scheduled for surgery in an ambulatory setting. Although the advantages of ambulatory surgery are well described [11], risk factors for unanticipated admission following ambulatory surgical procedures are poorly characterized [16, 23, 24]. Our study, which identified CP as a risk factor for unanticipated admission, is both novel and timely, given the increased emphasis on unanticipated admission as a crucial quality measure of the ambulatory surgical care process [25].
Previous literature has identified few risks associated with unanticipated admission including, ASA status ≥ 3, age < 2 years [23], and surgical specialty (with orthopedics, urology, and general surgery posing the greatest risk) [26]. In addition, our group recently demonstrated that obstructive sleep apnea is a risk factor for unanticipated admission following non-otolaryngologic surgery [16]. Other patient and surgical factors associated with unanticipated admission following ambulatory surgery have been poorly elucidated and remain an area in need of additional investigation. Thus, our study fills an important knowledge gap, demonstrating that children with CP require more careful examination during disposition planning following ambulatory surgery.
Although we do not have information on the antecedent events leading to unanticipated admission in our study population, other studies have demonstrated increased surgical complications in patients with CP. A recent study by Skertich et al. found that patients with CP undergoing an appendectomy or cholecystectomy incurred more postoperative adverse events, required more open procedures, and had longer hospital stays compared to the general population [27]. The ability to manage children with CP in the ambulatory surgical setting hinges on the anticipation of the challenges their comorbidities commonly present. For example, patients with CP often have chronic pain and sensory processing disorders and the presence of intellectual disabilities may limit their ability to communicate discomfort. This makes perioperative pain control challenging; unanticipated admission can result from poorly controlled pain [28-30]. Patients with CP commonly have chronic pulmonary diseases related to recurrent aspiration, impaired cough and airway clearance, scoliosis, chest wall deformity, upper airway obstruction, and respiratory muscle discoordination [31]. These may predispose the patient to perioperative respiratory events such as hypoxia and respiratory distress [32, 33]. Similarly, patients with CP may have GERD and excess salivation, both of which increase the risk of aspiration in the perioperative period [34]. Finally, it is estimated that 15%–55% of children with CP have comorbid epilepsy [35]. Epilepsy in the general population is known to carry a significantly higher risk of postoperative complications [36]. Consonant with these previous studies, we found that epilepsy increased the risk of unanticipated admission in children with CP by 16%. Given the significant degree of overlap of CP comorbidities with risk factors for adverse surgical outcomes and unanticipated admissions, evaluation of each CP patient and their individual clinical picture is needed to determine their appropriate operative setting. Future studies should focus on identifying the adverse events and perioperative challenges that contribute to unanticipated postoperative admissions in children with CP. Until we gain a more nuanced understanding of the factors that lead to these unanticipated admissions, recognizing the increased risk in this patient population can help surgeons and anesthesiologists set realistic expectations with families, optimize perioperative workflows, and reduce hospital costs associated with unexpected admissions. Likewise, when scheduling ambulatory procedures for children with CP, physicians now have evidence to support advocating for insurance preapproval of a postoperative hospitalization for those who may require it.
4.1 Limitations
This study is limited by the lack of granularity that is characteristic of large administrative database analyses. For example, we cannot determine individual hospital practices and disposition planning norms, patient complexity or American Society of Anesthesiologists Physical Status score, procedural timing or duration, causes of the unanticipated admission or the patient's hospital course. In our analysis, we defined unanticipated admission as a patient requiring a postoperative hospital stay greater than or equal to 1 day. However, this does not capture patients who may have needed an unanticipated admission but were discharged before 24 h. We also did not capture patients who were discharged home as planned after their procedure but presented less than 24 h later requiring readmission. Additionally, CP is a heterogeneous diagnosis and patients with CP are affected by varying degrees of motor impairment and comorbid disease burden. We acknowledge the potential limitations associated with the use of ICD-9 and ICD-10 codes in accurately identifying CP cases. These codes are often specific but not sensitive, and individuals identified by these codes may have more severe disease manifestations. Unfortunately, the PHIS contains no information on CP severity. To mitigate these limitations in future research, we suggest the development and application of more comprehensive validation studies for ICD codes used in identifying CP. Another potential limitation of our analysis is the adjustment for surgery based solely on specialty. Although this approach accounts for differences across broad categories, it may introduce residual confounding related to surgery type. Finally, children with CP often have a well-established rapport with their healthcare providers, which may influence the decision-making process about postoperative admissions. Our study aimed to highlight the increased risk of unanticipated admissions in this population, not to suggest that ambulatory surgery is inherently unsafe for children with CP. Indeed, the high proportion of successful ambulatory surgeries in our cohort of children with CP (over 90% were discharged as planned following surgery) supports the feasibility and popularity of ambulatory procedures with appropriate preoperative planning and individualized care. The possibility that clinicians might be more cautious with CP patients, leading to a higher admission rate, underscores the need for personalized discharge criteria and shared decision-making among the medical team, family, and patient.
5 Conclusion
Although CP is not a contraindication for ambulatory surgery in children, it is significantly associated with unanticipated hospital admissions. This underscores a need for careful preoperative clinical site of care selection in this vulnerable patient population.
Author Contributions
Ms. Nina Shamansky helped with the idea conception, study design, interpretation of the data, critical review of literature, initial writing of the manuscript, and manuscript revision. Dr. Christian Mpody helped with the idea conception, study design, data acquisition and analysis, interpretation of the data, manuscript preparation, and manuscript revision. Dr. Olubukola O. Nafiu helped with idea conception, data acquisition and analyses, critical review of literature, and manuscript preparation and revision. Dr. Joseph D. Tobias helped with interpretation of the data, critical review of literature, manuscript preparation, and manuscript revision. Dr. Brittany L. Willer helped with the idea conception, study design, interpretation of the data, critical review of literature, initial writing of the manuscript, and manuscript revision. All authors approved the final manuscript.
Conflicts of Interest
The authors declare no conflicts of interest.
Open Research
Data Availability Statement
Data sharing is not applicable to this article as no new data were created or analyzed in this study.