1 Introduction

Spastic hip dysplasia commonly affects pediatric patients with cerebral palsy, as it occurs in approximately 35% of these children.¹ Additionally, hip dysplasia risk is correlated with the child's Gross Motor Function Classification System (GMFCS) level.^1,2 Children with GMFCS scores of IV and V have a 69% and 89% chance of developing hip dysplasia respectively.³ Surgical intervention for hip dysplasia can be categorized into preventive, reconstructive, and salvage procedures. Multiple factors including age of the patient, hip migration percentage (MP), and acetabular index help physicians determine the type of surgical procedure that should be performed. In general, an osseous reconstruction is indicated for children with cerebral palsy with an MP greater than 40%, while salvage procedures are indicated when a hip is “unreconstructible” or when a reconstruction procedure is unlikely to succeed.^4–8

Our study is primarily concerned with the outcomes and complications of reconstructive and salvage procedures for the hip. Previous literature has reported complication rates ranging from 10 to 65% for reconstructive procedures^8–10 and 19–33% for salvage procedures.^11,12 These relatively high rates are understandable due to cerebral palsy's association with multiple medical comorbidities such as seizure disorders, gastrointestinal and nutritional problems, and respiratory issues.¹⁰ These comorbidities may require medical interventions such as a gastrostomy tube (G-tube) or a tracheostomy, which can increase the risk of complications occurring perioperatively.¹³ Some of the most common complications of these hip procedures include heterotopic ossification (16%), osteonecrosis (1–11%), and proximal migration of the femur (8%).^9,11,14 Additionally, insufficiency fractures and decubitus ulcers are typical complications of postoperative casting procedures.^6,15 Dysplasia is also more likely to recur in patients with higher preoperative MP and GMFCS ratings as well as older patients.⁷ These operative complications and preexisting comorbidities make caring for children with cerebral palsy a complex task and require many separate appointments with multiple medical specialties for a single child.

Much of the current literature investigating hip surgeries in pediatric cerebral palsy consists of studies from a single institution and variable follow-up times and sample sizes. For this reason, there is a relative lack of information regarding readmission rates, common reasons for readmissions, and specific risk factors or comorbidities impacting readmissions on a larger scale. Our study hopes to expand on these limitations by utilizing the Nationwide Readmissions Database (NRD). The NRD has multiple inherent advantages such as the fact it contains data from many care centers and both inpatient and post-discharge outcomes within 90 days. The NRD additionally has information on readmission rates and cost analysis which are not as thoroughly covered in the current literature. Therefore, the primary goal of this research is to generate a more comprehensive understanding of how children with cerebral palsy are affected by hip surgery on a national level with respect to readmission rates and specific associated risk factors for readmission as well as prolonged length of hospital stay.

2 Methods

3 Results

The NRD showed a total of 36,308 pediatric patients with cerebral palsy in the United States from 2014 to 2018. After exclusion of patients admitted non-electively and during the fourth quarter for readmission analysis purposes, 1225 patients were included in the cohort for analysis. Exact selection process according to inclusion and exclusion criteria can be seen in Fig. 1. This population was, on average, 9.3 ± 3.8 years old and 42.8% female. The majority of patients were covered by Medicaid insurance (58.4%) and treated at private, not-for-profit (82.1%), metropolitan teaching hospitals (94.0%). See Table 1 for entire patient demographics and treatment site information. The most common comorbidities noted preoperatively in this population included chronic pulmonary disease (10.5%), fluid and electrolyte disorders (6.9%), and cardiac arrhythmias (6.6%). See Table 2 for all reported preoperative comorbidities.

Close

Fig. 1 Selection criteria of patient cohorts.

Export to PPT

Additionally, 174 patients (14.2%) required readmission within 90 days of surgery. Compared to patients who were not readmitted within 90 days, readmitted patients were similar in age (8.9 ± 3.8 years vs. 9.4 ± 3.8 years, p = 0.09) and sex (44.8% female vs. 36.4% female, p = 0.61), but more likely to be covered by public insurance (69.5% vs. 56.6%, p < 0.01). Additionally, readmitted patients were more likely to have certain comorbidities including cardiac arrhythmias (13.8% vs. 5.4%, p < 0.01), coagulopathy (5.2% vs. 1.2%, p < 0.01), chronic pulmonary disease (19.0% vs. 9.0%, p < 0.01), hypertension (3.4% vs. 1.1%, p = 0.02), iron deficiency anemia (3.5% vs. 0.5%, p < 0.01), depression (2.9% vs. 0.4%, p < 0.01), fluid and electrolyte disorders (21.3% vs. 4.4%, p < 0.01), and weight loss (8.6% vs. 2.9%, p < 0.01). The most common diagnoses upon readmission included gastrostomy status (17.8%), gastro-esophageal reflux disease without esophagitis (13.8%), psychological development disorders (12.6%), and epilepsy (10.9%). For the list of the most common medical diagnoses upon readmission see Table 3.

The mortality rate of the inpatient population was 0.1%. The majority of patients had a routine disposition at discharge (86.6%) and the average LOS was 4.5 ± 7.0 days. Additionally, 26.3% of patients had a prolonged LOS (≥5 days). The average amount of hospital charges incurred was $82,236.19 ± $86,396.51 where hospital stays less than 5 days cost on average $64,514.51 ± $36,191.14 and stays of 5 days or greater more than doubled to $131,757.52 ± $146,057.01. There was no difference in either of these between readmitted and non-readmitted patients for overall average cost (p = 0.56), average cost for prolonged stays (p = 0.35), and average cost for short stays (p = 0.99). The all-cause complication rate, or the proportion of patients who sustained at least one complication throughout their treatment, was 5.5% overall. The most common medical complications included acute respiratory failure (2.0%), pneumonia (1.6%), urinary tract infection (0.9%), sepsis (0.7%), and paralytic ileus (0.6%). Patients requiring readmission had significantly more cases of acute respiratory failure than patients who were not readmitted (4.6% vs. 1.6%, p = 0.02). The most common surgical complication was dehiscence (0.2%). Other surgical procedures performed in conjunction with the hip procedures included blood transfusion (2.9%) and mechanical ventilation (0.8%). See Table 4 for complete results on LOS and charges and Table 5 for inpatient complications sustained during admission.

Multivariate analysis identified several factors associated with extended LOS and 90-day readmission. Patients with a prolonged hospital stay were more likely to have a nonroutine discharge (odds ratio [OR] 2.26, p < 0.01), an inpatient medical complication (OR 3.35, p < 0.01), an inpatient surgical complication (OR 14.67, p = 0.01), a cardiac arrhythmia (OR 2.63, p < 0.01), and iron deficiency anemia (OR 4.52, p = 0.02). Patients readmitted within 90 days were more likely to have Medicaid status (OR 1.67, p < 0.01), medical complications (OR 2.68, p < 0.01), cardiac arrhythmias (OR 1.81, p = 0.04), chronic pulmonary disease (OR 2.11, p < 0.01), iron deficiency anemia (OR 6.04, p < 0.01), fluid and electrolyte disorders (OR 3.72, p < 0.01), and weight loss (OR 2.42, p = 0.02). Finally, patients sustaining an inpatient medical complication were more likely to have Medicaid status (OR 2.20, p = 0.01), being treated at a non-metropolitan hospital (OR 11.14, p = 0.03), and a nonroutine discharge (OR 1.97, p = 0.03). Multivariate analysis regarding prolonged stay, readmission within 90 days, and inpatient medical complications can be fully seen in Tables 6 and 7.

4 Discussion

Spastic hip dysplasia in children with cerebral palsy worsens as they grow, which will likely negatively affect quality of life (QOL) if left untreated. Conservative treatments such as physical therapy or botulinum injections may slightly delay reconstructive surgery, but literature has not shown that these interventions will prevent surgery entirely.⁷ Additionally, reconstructive surgery has been shown to improve QOL and effectively treat hip dysplasia in cerebral palsy pediatric patients.¹⁸ Therefore, surgical intervention will likely continue to be the primary treatment option for this population. However, these necessary reconstructive and salvage procedures can result in various complications and long hospital stays. Thus, it is critical for healthcare teams to be mindful of these risks to treat their patients most effectively and efficiently. This study aimed to expand upon the current literature by establishing national complication and readmission rates and identifying specific risk factors and complications affecting pediatric cerebral palsy patients undergoing surgical hip procedures.

It is well understood that cerebral palsy is a highly complex medical condition and has a significant impact on the patient's postoperative course. Literature has shown that children with cerebral palsy are more likely to experience postoperative complications compared to other children.¹⁹ However, the majority of research regarding readmissions due to specific surgeries in this population involves spinal deformity surgery, which has shown readmission rates ranging from 16 to 18%.^20,21 One study investigating bilateral hip reconstruction found a readmission rate of 9.2% in this population.²² We expand upon the literature by establishing a national readmission rate after hip surgery of 14.2% and identifying the most common reasons for readmission including gastrostomy status (17.8%), gastro-esophageal reflux (13.8%), and epilepsy (10.9%). Postoperative care teams should emphasize these possible events when communicating with the patient's caregivers and educate them on relevant early signs and symptoms that their child may require readmission. This may help caregivers identify any potential causes for readmission earlier in the disease process, allowing them to seek medical guidance in a more time effective manner.

Patients with cerebral palsy also commonly have pre-existing comorbidities that impact their medical care and must be taken into account by their healthcare providers. This study identifies specific comorbidities including cardiac arrhythmias (LOS OR: 2.63; Readmission OR: 1.81) and iron deficiency anemia (LOS OR: 4.52; Readmission OR: 6.04) increased the odds for children with cerebral palsy to have both an elongated LOS postoperatively and also be readmitted within 90 days. Additionally, patients with chronic pulmonary disease (OR: 2.11), fluid/electrolyte disorders (OR: 3.72), and significant weight loss (OR: 2.42) have an increased chance of being readmitted. These comorbidities should be emphasized on preoperative assessment when determining if a patient with cerebral palsy and at least one of the above comorbidities is safe for hip surgery. Care teams should take the necessary steps to place patients in the best possible position to handle surgery, such as ensuring weight stabilization before approving a patient for surgery. Previous research has shown that preoperative multidisciplinary screening and shared decision making improved functional outcome in upper-extremity surgeries for patients with cerebral palsy.²³ This multidisciplinary care model allows for complex patients to see multiple providers and allows these providers to directly interact and coordinate their plan of care, while also allowing patients and their families to more efficiently access care.^24,25 Along these lines, encouraging shared decision making between providers and caregivers might allow for more personalized care for children with cerebral palsy, ensuring specific comorbidities are considered before, during, and after hip surgery. This allows care teams to make complex medical decisions with families regarding their child's health and also potentially reduces readmission rates and improves patient outcomes.

As previously discussed, children with complex medical conditions such as cerebral palsy are inherently more likely to encounter complications surrounding medical procedures.³ This study found an overall medical complication rate (5.0%) and surgical complication rate (0.5%) that were lower than previous literature reports, potentially due to the use of the NRD and its coding procedures. While heterotopic ossification and osteonecrosis have previously been reported as some of the most common complications of these hip surgeries,^9,11 there were no instances of these in our population. Instead, we found respiratory complications, such as acute respiratory failure (2.0%), pneumonia (1.6%), and urinary tract infections (0.9%) to be most prevalent. This study does identify multiple non-medical factors that increase their likelihood of sustaining an inpatient medical complication. These factors included having Medicaid insurance (OR: 2.20) and being treated at a non-metropolitan hospital (OR: 11.14). Sustaining any medical complications additionally increases the likelihood of having an elongated LOS (OR: 3.35) and a readmission within 90 days (OR: 2.68) making this somewhat of a vicious cycle for patients that fall into these categories. It has also been shown that outcomes are improved and fewer reconstructive surgeries are required when surgery is performed earlier in the disease process, reinforcing the importance of early hip surveillance for children with cerebral palsy.²⁶ Patients seen at non-metropolitan hospitals and those with public insurance may be more likely to have more barriers to access care, potentially delaying a necessary surgery and leading to these noted differences in outcomes. Identifying patients in these categories preoperatively and providing them with additional support may help address pertinent complex healthcare and social risk factors to improve their care. Forms of additional support may feature shared decision making or case conferences dedicated to holistically caring for patients.

This study required the use of a national database and is thus subject to the limitations that are associated with it. While cerebral palsy is a heterogeneous condition, the NRD does not stratify by type of cerebral palsy and is unable to portray severity of both cerebral palsy (GMFCS level) and underlying comorbidities. Therefore, we are unable to relate these factors, as well as intraoperative characteristics such as estimated blood loss, to our primary variables of readmission and length of stay. Additionally, we relied on accurate ICD-9 and ICD-10 coding for the correct identification of pediatric cerebral palsy patients requiring a reconstructive hip procedure. The NRD only contains post-discharge complications that have been coded and is limited to readmissions within 90 days and to hospitals associated with the NRD. Therefore, our reported complication rates potentially underestimate the true prevalence of complications following hip reconstructive surgery in cerebral palsy patients. Additionally, certain complications such as heterotopic ossification may present to outpatient visits for follow-up rather than readmission or require readmission after 90 days, resulting in decreased complication rates in this study. Finally, while the NRD typically recommends only analyzing patient variables with 11 or more qualifying patients, we chose to incorporate these variables due to a smaller group of interest. A study that can analyze longer-term complications and readmission characteristics would be beneficial in understanding how surgical hip procedures affect children with cerebral palsy on a larger time scale.

5 Conclusion

This study uses a nationwide database to evaluate key factors impacting elongated LOS, readmissions, and treatment complications for children with cerebral palsy undergoing reconstructive and salvage hip surgeries. Healthcare teams should identify patients with specific comorbidities that place them at risk for elongated LOS and postoperative readmission, including cardiac arrhythmias and iron deficiency anemia. These may be effective steps towards this goal and improving the overall quality of care for this patient population. One possible method of enacting these changes is implementing a multidisciplinary, coordinated model of care for patients with cerebral palsy that can not only identify patients in need of hip surgery at an earlier time, but also more effectively determine the appropriate timing for surgery in these patients for optimal outcomes.

Funding/sponsorship

This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors.

Informed consent

n/a.

Institutional ethical committee approval

n/a.

Authors contribution

Author 1 – conceptualization, data curation, formal analysis, investigation, methodology, writing original draft, review & editing.

Author 2 – data curation, formal analysis, investigation, methodology, review & editing.

Author 3 – conceptualization, data curation, investigation, methodology, writing original draft.

Author 4 – data curation, formal analysis, investigation, methodology, review & editing.

Author 5 – Investigation, methodology, project administration, resources, review & editing.

Author 6 – Investigation, methodology, project administration, resources, review & editing.

Author 7 – conceptualization, investigation, methodology, project administration, resources, review & editing, validation, visualization.