Do Pediatric Hospitals Improve Operative Efficiency?

Abstract

Background

In recent years there has been a push towards developing free standing pediatric facilities to provide care specifically towards pediatric patients. The purpose of this study was to determine if moving pediatric cases from a general hospital to a dedicated pediatric facility improved the quality and efficiency of surgical procedures.

Methods

A retrospective review of pediatric patients undergoing posterior spinal fusion (PSF) was performed. All procedures were performed by one orthopaedic surgeon (SLW) from 2015 to 2019. The procedures were performed at a general hospital (GH) the first two years, and at a pediatric hospital (PH) the subsequent years. Data extracted included patient sex, age, and procedure type as well as procedure duration, operative turnover time, hospital length of stay, transfusion requirements, and operative delay. Exclusively pediatric adolescent idiopathic scoliosis (AIS) patients undergoing PSF were included due to the high volume and consistent surgical procedures therefore limiting confounding variables.

Results

A total of five hundred PSF pediatric procedures were performed during the time period. After excluding non-adolescent idiopathic scoliosis cases, a total of 208 procedures were reviewed (105 at GH; 103 at PH). There was no statistical difference between the groups in regards to operative time (GH: 200 min, PH: 200 min; p=0.91), room turnover time (GH: 38 min, PH: 38 min; p=0.801), or rate of transfusion (GH: 20% PH: 30%; p=0.09). Length of stay was significantly shorter in the PH cohort compared to the GH cohort (4.35 vs. 3.84 days, p=0.0001). However, a smaller proportion of cases at the PH started on time compared to the GH (34% vs. 58%; p=0.0005).

Conclusion

Overall, this study demonstrated that AIS procedures at the PH did show a statistically significant reduction in hospital length of stay. However, timely start of the procedure was less likely at this particular facility.

Level of Evidence: III

Introduction

There has been a long history of providing medical care specific to children in the United States and Canada. In 1855 the Children’s Hospital of Philadelphia developed a facility dedicated to children. Since then this trend has rapidly spread throughout the developed world.¹ Over the next century, health care continued to evolve and previously high-volume childhood ailments that resulted in high childhood mortality and handicap decreased. In light of this changing healthcare landscape, the role of freestanding pediatric facilities started to be called into question. This, coupled with the increased cost of healthcare, caused many health care organizations to seek opportunities for cost saving initiatives. As a result, many previously free-standing children’s hospitals either closed completely or merged with adult based hospitals.¹

While this occurred on a wide scale, a number of pediatric hospitals have been able to thrive, both economically and with providing high quality patient care.² These pediatric hospitals have typically been associated with academic institutions as well as continued to maintain strong connections with their general hospital counterparts.¹

Orthopaedic surgeries are the second most common pediatric inpatient procedures and dedicated hospital facilities and operating room personnel geared specifically to this age demographic have been shown to be beneficial.³ The injuries and illnesses encountered by this population are unique, requiring technically-demanding surgical procedures and carefully planned post-operative care. In light of this, a specialized health care team familiar with the particular needs and post-operative protocols for this population can maximize results while minimizing complications.³

Adolescent idiopathic scoliosis (AIS) is one of the most common clinical conditions treated by pediatric orthopaedic surgeons. This condition is characterized by a coronal curvature of the spine measuring greater than 10 degrees utilizing the Cobb angle^4,5 in the absence of underlying structural or syndromic causes. It has been reported to affect roughly 3% of the general population from age 10-16 and more commonly affects females than males.^6,7 Of this number, only 0.3% to 0.5% of those affected will have a curve greater than 20 degrees at which point treatment would be recommended.⁸ These treatments have typically consisted of bracing treatment for curve angles from 20-40 degrees to help prevent curve progression in skeletally immature individuals. Surgical intervention has typically been reserved for curves measuring greater than 45 degrees.⁹ Surgical techniques have certainly evolved over the years, but the current gold standard of treatment is a posterior spinal fusion utilizing a pedicle screw construct and connecting rods. These surgical interventions are undertaken to avoid the long term sequelae of untreated AIS which can include curve progression, back pain, pulmonary restrictive effects, and psychosocial issues.^9,10 In the United States, roughly thirty thousand scoliosis surgeries are performed annually.¹¹ These are typically very involved and intricate procedures with surgical case times very much surgeon dependent. Similarly, post-operative length of stay is certainly variable is often quite lengthy.

While previous studies have examined the beneficial impact of having a dedicated OR team in decreasing procedure length and minimizing blood loss in pediatric AIS cases,¹² no studies were found investigating these differences between general and pediatric hospitals settings. The purpose of this study was to determine if AIS corrective procedures performed at a pediatric specific hospital affects operating room variables by investigating procedure duration, hospital length of stay, transfusion requirements, and operative cases starting time and comparing these to previously obtained data from the general hospital. The second aim was to evaluate the difference in cost efficiency between the two facilities.

From an experimental set up perspective, the senior staff surgeon (SLW) was constant between the two facilities and was well familiarized with the operative procedure. There were no changes in operative technique, instrumentation, or preoperative work up. A third year orthopaedic resident continued to serve in the first assist role and was responsible for instrumenting his/her side of the spine. From additional staffing perspective, a dedicated AIS scrub nurse and circulator continued to be utilized. The only difference between the two facilities was the usage at the pediatric hospital of a preoperative timeout requiring representation from the surgical and anesthesia team along with the patient’s family prior to transport to the operating room. Based upon these operative constants, we hypothesized that there would be no difference in operative variables between the cases completed at the general hospital to those completed at the stand-alone pediatric facility.

Methods

After institutional review board approval, we queried the electronic medical record database for all procedures identified by CPT codes 22843 and 22844 (Posterior spinal instrumentation including less than 13 vertebral segments and posterior spinal instrumentation including 13 or more vertebral segments) during the years 2015-2019, inclusive. This interval included the 2 years prior to, and 2 years after, transition of all pediatric procedures from a general hospital to a stand-alone children’s hospital associated with the same academic medical center. All procedures in the sample were performed by one pediatric orthopedic surgeon with greater than 30 years’ experience in the same institution. This initial data search resulted in over 500 results. This scope was further narrowed by applying exclusion criteria that limited results to solely patients aged 10 years or older and without syndromic or structural causes of scoliosis. This resulted in 105 cases in the GH cohort and 103 cases in the PH cohort. For each procedure we extracted the patient age and sex, extent of the fusion (<13 levels or >/=13 levels) as well as whether the case started on time, length of time from patient in room to procedure start (minutes), length of the procedure from incision to final closure (minutes), turnover time measured from patient out of room to next patient in room (minutes), hospital length of stay (days), transfusion requirements in the immediate postoperative period as recorded in the surgical and post-operative clinic notes, pediatric specific anesthesiologist staffing, and certified registered nurse anesthetist (CRNA) staffing.

The distributions of the variables within each sample were characterized using the median and interquartile range for ordinal-level variables, and proportions for nominal-level variables. The Wilcoxon two-sample test was used to compare continuous variables including age, operative time, turnover time, blood loss, and length of stay. Fisher’s exact and Pearson chi-square tests were used to evaluate differences in sex and occurrence of complications. Stata (version 15.1; Stata Corp, College Station, TX, USA) was used for all statistical analysis; differences associated with a p-value <0.05 were considered to statistically significant.

Results

A total of 500 patients underwent PSF by the surgeon during the 4-year time frame. Of these, 292 were excluded due to age less than 10, or a diagnosis other than AIS. This resulted in a final cohort of 208 patients, with 105 cases performed at the GH and 103 at the PH. Females made up 84% of the sample at the GH compared to 76% at the PH (p=0.13 The median operative time (incision to closure in minutes) was 193 min (IQR=174–215 min) at the GH and 196 min (IQR=172–213 min) at the PH (p=0.91). There was likewise no statistical difference in median turnover times (time elapsed between one patient leaving the OR and the next entering) between the facilities (36.5 min (IQR=32.5–44 min) and 37.5 min (IQR=33–39 min) at the GH and PH, respectively, p=0.99). However, the mean in-room-time to case-start time was 5.6 minutes faster at the PH compared to the GH (mean 48.9 minutes GH vs. 54.5 minutes PH, p<0.0001). There was an average delay of cases starting on time of 42 minutes at the pediatric facility compared to 13 minutes at the general hospital (p<0.0001), resulting in a lower proportion of on-time starts (34% vs. 58%) as measured by cases starting within 5 minutes of their scheduled start time. Variances in type of anesthesia staff utilized were also analyzed, with a statistically significant difference in CRNA utilization (91% PH vs. 61% GH) and pediatric-trained anesthesia staff (99% PH vs. 84% GH) when comparing the two facilities (p<0.0001). There was also a slightly higher incidence of longer fusion constructs in the children’s hospital with 84% of procedures being >13 levels as compared to 71% at the general hospital (p=0.04). Similarly, there was a slightly higher incidence of transfusion requirement in the pediatric hospital with 30% requiring a transfusion compared to 20% in the general hospital, however these findings were not significant (p=0.09). In terms of hospital length of stay, there was a statistically significant difference with the designated children’s hospital length of stay being 0.5 days shorter on average (p=0.0001) at 3.7 days compared to 4.2. These results are summarized in Figure 1 and Figure 2.

Figure 1.

Figure 2.

Discussion

This study provides valuable data on specific indicators of OR efficiency and clinical outcomes, allowing us to target areas for improvement at the pediatric specific facility. Specifically, migration to the PH was associated with cases typically starting thirty minutes later than similar procedures at the GH. On the positive side, LOS was decreased by approximately 0.5 days after migration to the pediatric hospital but given the national trends toward shortened hospital stays, this was likely to occur without a change in facility. Additionally, a higher percentage of CRNA utilization and pediatric board certified anesthesiologists was also noted following the migration to the pediatric hospital. This was relatively unsurprising to the authors given the increasing utilization of CRNA’s in the surgical realm and the pediatric-specific facility.

Recent literature has focused on operative efficiency from several different perspectives. Some studies have evaluated the role of consistent OR personnel, use of dedicated operating rooms for specific procedures, and incorporating surgical co-management services in improving procedural efficiency.^13-15 To the best of our knowledge, procedural efficiency as a function of type of facility – PH vs GH - has not been evaluated to date.

This study suggests that at our institution, utilizing the same staff surgeon with resident assistance model, procedural length was unaffected by the move to the PH. This supports the authors’ original hypothesis that procedure length is a function of the staff surgeon’s strong familiarity with adolescent idiopathic scoliosis and posterior spinal fusion. It was of keen interest to the authors that there was a statistically significant difference in procedures starting on time. With migration to the dedicated pediatric facility, there was an average twenty-nine minute longer delay relative to that at the general hospital. While it is impossible to isolate one reason for this, it is likely largely attributable to a mandated pre-procedural time out conducted in the preoperative holding area. This time out is unique to the pediatric hospital and requires that all members of the surgical team be present prior to transporting the patient to the operative room suite. This requirement was implemented in an attempt to minimize surgical errors and to ensure that parents were fully informed of the procedure about to be undertaken. We agree with the intent, however, due to the various schedules and responsibilities delays delaying transport until all members are present does This process has been largely successful at this aim, but with stipulating that all members of the surgical team must be present prior to patient transport delays in patient transport have been abundant.

Length of stay decreased from a median of 4.3 at the GH to to 3.8 days after the move to the PH. This 0.5 day savings in patient length of stay represents a significant cost savings to the patient and a significant cost savings to the hospital when under bundled care plans where a flat fee is reimbursed for all services provided. In addition to this, earlier discharge also allows for increased patient bed availability. This length of stay decrease also compares favorably to the national average for length of stay from 2017 of 4.1 days for AIS procedures in patients age 10-17.¹⁶

We cannot assume that the experience of other hospital systems would be similar to ours due to the highly individualized nature of each hospital system. Our situation was very unique in that our institution predominately utilizes one surgeon for all AIS cases and his operative case volume is very high. This study does provide insight into the importance of ensuring pre-operative checklists are effective at eliminating preventable complications and improving communication between the patient, family and surgical team, without routinely causing unnecessary delays. Additionally, this study provides some evidence that the familiarity of a surgeon with a particular procedure is more important in operative efficiency than dedicated facilities for particular age groups.

There are limitations to be considered when evaluating the results of this study. Data was collected retrospectively, and we cannot make any claims about the accuracy of the times recorded by the OR personnel. In addition, this study was very specific with regards to our particular institution and is likely not generalizable to other facilities. With these problems being known, this study does provide valuable information for our institution regarding targets for efficiency improvement moving forward. It demonstrated that while operative procedure length may not have varied with migration to a new facility, this migration did result in a larger number of cases not starting on time and that the preoperative timeout prior to transition to the operating room was likely responsible for this.