Published online April 1, 2008
PEDIATRICS Vol. 121 No. 4 April 2008, pp. e906-e911 (doi:10.1542/peds.2007-2089)
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow P3Rs: Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when P3Rs are posted
Right arrow Alert me if a correction is posted
Services
Right arrow E-mail this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My File Cabinet
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via CrossRef
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Gregory, C. J.
Right arrow Articles by Marcin, J. P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Gregory, C. J.
Right arrow Articles by Marcin, J. P.
Related Collections
Right arrow Emergency Medicine

ARTICLE

Comparison of Critically Ill and Injured Children Transferred From Referring Hospitals Versus In-House Admissions

Christopher J. Gregory, MDa, Farid Nasrollahzadeh, MDa, Madan Dharmar, MBBSa, Kourosh Parsapour, MDa and James P. Marcin, MD, MPHb

a Department of Pediatrics, University of California Davis Children's Hospital
b University of California Davis Center for Health Services Research in Primary Care, University of California, Davis, California


   ABSTRACT
TOP
 ABSTRACT
METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
OBJECTIVE. The purpose of this work was to compare the outcomes, severity of illness, and resource use of patients transferred to PICUs from outside hospitals to patients admitted from within the same hospital.

METHODS. We conducted a secondary analysis of patients from the 20 US PICUs in the most recent Pediatric Intensive Care Unit Evaluations Software Recalibration Database on a total of 13017 emergent PICU admissions between January 2001 and January 2006. Dependent variables were PICU resource use and risk-adjusted mortality. The main independent variable was the PICU admission source: patients transferred from referring emergency departments and inpatient wards versus in-house admissions from the same hospitals' emergency departments and inpatient ward.

RESULTS. Patients admitted from referring emergency departments had higher use of vasoactive infusions (7.31% vs 5.23%) and mechanical ventilation (33.45% vs 23.6%) than same-hospital emergency department admissions. Compared with in-house ward admissions, patients transferred from referring inpatient wards had higher mechanical ventilation rates (45.05% vs 28.56%) and PICU lengths of stay (8.0 vs 6.7 days).

CONCLUSIONS. On average, children admitted to a cohort of US PICUs from referring hospitals were more ill and required more intensive care resources than patients admitted to the same PICUs from within the institution. Hospital-level differences in PICU efficiency and severity of illness were highly variable. These data highlight the need for standardized PICU admission criteria to maximize hospital efficiency and suggest opportunities for earlier intervention and consultation by hospitals with PICU-level services to improve quality of care for critically ill children.

Key Words: hospitalized children • patient transfer • severity of illness • efficiency • outcome assessment • critical illness

Abbreviations: PRISM III—Pediatric Risk of Mortality III • PICUE—Pediatric Intensive Care Unit Evaluation • ED—emergency department • LOS—length of stay

Previous studies have demonstrated improved patient outcomes for pediatric patients cared for in dedicated PICUs.1–6 This has prompted organizations such as the Society of Critical Care Medicine and the American Academy of Pediatrics to issue recommendations for the regionalization of pediatric critical care services in the United States.7 The limited availability of PICU space and staff in such a system necessitates that only patients who absolutely require these services be referred and transported to regional PICUs. Efficient use of PICU resources is often measured by the proportion of patients within an ICU who have a high probability of death or require PICU-specific therapies, such as mechanical ventilation and vasoactive infusions.8,9 Whether referring hospitals are appropriately identifying and triaging critically ill children to hospitals with PICU capabilities has been questioned previously.10,11

Studies examining the severity of illness and outcome measures of patients transferred from outside facilities to regional PICUs are rare. A single-center study from the United Kingdom demonstrated that patients transferred to a regional PICU were usually considered appropriate admissions, as defined by either requiring PICU-specific therapies or having a predicted risk of mortality of >1% as determined by the Pediatric Risk of Mortality III (PRISM III) score.9 It is uncertain whether PICU resources are being as efficiently used by referring hospitals in diverse areas of the United States. These data are important in understanding current hospital use and limitations and in assisting clinicians and policy-makers to better manage limited PICU resources. The hypothesis that we sought to test in this study was that there were important differences in the severity of illness, use of PICU-specific therapies, and patient outcomes between children transferred from outlying hospitals and children admitted from within hospitals with PICU capabilities.


   METHODS
TOP
 ABSTRACT
 METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Data Source
To assess differences in patients' severity of illness, use of PICU-specific therapies, and patient outcomes, we analyzed data from the most current Pediatric Intensive Care Unit Evaluations (PICUEs) Software Recalibration Database. This database contains patient-level and hospital-level characteristics of a cohort of 24485 pediatric patients from 21 PICUs in the United States. One center was excluded for missing data needed to calculate PRISM III scores and data on the use of ICU-dependent technology, leaving 20 centers in the final database. This database was generated from sites that voluntarily participated in the PICUEs algorithm recalibration effort and that met PICUEs reliability standards.8 Details of the data collection procedures and exclusion criteria have been published previously.8 A minimum of 67 demographic, diagnostic, physiologic, laboratory, and outcome data points were collected for all of the patients within 24 hours after admission to the PICU. Data for analysis were collected between January 2001 and January 2006. Patients not included in the database were those who stayed in the PICU <2 hours, patients admitted in a state of continuous cardiopulmonary resuscitation who never achieved stable vital signs for ≥2 hours, and those patients who were admitted for recovery from procedures normally cared for in other hospital locations. All of the data included in the database met PICUE reliability standards, as assessed by data reabstraction and reliability testing.8,12

Patients and Inclusion Criteria
A priori, we limited our analysis to patients admitted to participating PICUs on an emergent basis. We believed that this would give the most representative comparison of interhospital and intrahospital admissions, because very few patients were admitted to PICUs from outside facilities on an elective basis.

All of the patients were included if they were ≤18 years of age at time of PICU admission and were admitted to the PICU from the emergency department (ED) or inpatient ward. Patients admitted from other locations in the hospital, such as the operating room or cardiac catheterization laboratory, were not included in the analysis. The primary dependent variables included were PICU efficiency as defined by Pollack et al5 (use of vasoactive medications within 24 hours of admission or use of mechanical ventilation within 24 hours of admission), PRISM III-calculated risk of mortality, and severity-adjusted mortality ratio. The primary independent variables for these analyses were hospital origin (outside hospital versus same hospital) and hospital location (ED and inpatient ward). Severity-adjusted mortality rates were calculated as the overall observed mortality rate of the patients meeting inclusion criteria multiplied by the observed-to-expected mortality rate ratio of the individual PICUs.

Statistics
Categoric variables were compared by using a {chi}2 test. We used Student's t test for univariate comparisons of continuous normally distributed variables and the Wilcoxon rank-sum test when the data were not normally distributed. Comparisons of proportions were compared by using a test of proportions. We conducted all of the statistical analyses by using Stata 8.0 (Stata Corp, College Station, TX). We considered P values of <.05 to indicate statistical significance. The study was approved by the human subjects review committee at University of California Davis.


   RESULTS
TOP
 ABSTRACT
 METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
There were 14728 patients admitted on an emergent basis from either an ED or inpatient ward to the 21 PICUs in the database. A total of 1711 patients (11.6%) were excluded for missing data, leaving 13017 patients in the final analysis. The database was missing data for use of vasoactive medications in 1543 patients (10.5%). The use of mechanical ventilation was missing for 575 patients (3.9%), the hospital of origin was missing for 131 patients (0.89%), and 38 patients (0.25%) were missing data needed to calculate PRISM III scores. A summary of the demographic characteristics of the participating centers is listed in Table 1.


View this table:
[in this window]
[in a new window]

  		TABLE 1 Demographic Description of 20 Participating PICUs

 
As shown in Table 2, the most common primary PICU-admitting diagnoses from referring and same hospitals were trauma (20.7% and 13.4%, respectively) and bronchiolitis or pneumonia (14.6% and 18.6%, respectively). There were proportionally more admissions from outside hospitals for drug overdoses, seizures, and congenital heart disease when compared with intrahospital admissions. Patients were more frequently admitted from within the same institution for reactive airway disease and trauma. Comparing patients admitted from referring hospitals with patients admitted from within the same facility, there were no differences between groups in terms of gender, but transferred patients were younger (Table 3).


View this table:
[in this window]
[in a new window]

  		TABLE 2 Primary Admission Diagnoses, Stratified by Admission Source

 

View this table:
[in this window]
[in a new window]

  		TABLE 3 Patient Characteristics, Severity of Illness, and Use of PICU-Specific Therapies (n = 13 017), Stratified by Admission Source

 
ED Admissions
Patients admitted from the EDs of referring hospitals had higher rates of mechanical ventilation (33.5% vs 23.6%; P < .01) and vasoactive infusion (7.3% vs 5.2%; P < .01) on the first day of ICU admission than patients admitted from their own hospitals' EDs (Table 3). Overall, PICU efficiency was higher for patients admitted from outside EDs (34.9%) than for patients admitted from the EDs of the same institution (24.7%). Referring ED admissions had a trend toward higher mean PRISM III scores (4.40 vs 4.16; P = .06) compared with same-hospital ED admissions. There was no difference in severity-adjusted mortality rates between patients admitted from referral hospital EDs and those admitted from within the same hospital EDs.

Inpatient Ward Admissions
Patients admitted from the inpatient wards of referring hospitals had higher rates of mechanical ventilation (45.1% vs 28.6%; P < .01) but similar rates of vasoactive infusion use (11.3% vs 9.5%; P = .12) and PRISM III scores (5.95 vs 5.62; P = .22) than patients admitted from within the same hospital (Table 3). Overall PICU efficiency was higher for patients admitted from outside inpatient wards (47.7%) than for patients admitted from the inpatient wards of the same hospital (33.04%). Severity-adjusted mortality rates were similar between patient cohorts. PICU length of stay (LOS) was significantly longer for patients transferred from the inpatient wards of outside hospitals compared with in-house admissions (8.0 vs 6.7 days; P = .01).

ED Versus Inpatient Ward Admissions
Patients admitted from inpatient wards, from both referring hospitals and in-house admissions, had higher PICU efficiency rates than patients admitted from the ED (37.8% vs 28.3%; P < .01). Similarly, PRISM III scores were higher for patients admitted from the inpatient wards compared with the ED from either outside hospitals (5.95 vs 4.40; P < .01) or within the same hospital (5.62 vs 4.16; P < .01).

Individual PICUs
Table 4 shows the individual descriptions of the 20 participating PICUs. There was significant variation among centers in the proportion of patients transferred from outside facilities. Similarly, there was significant variation among centers in their use of mechanical ventilation and vasoactive infusions, PICU efficiency, mean PRISM III score, and severity-adjusted mortality rate.


View this table:
[in this window]
[in a new window]

  		TABLE 4 Outcomes Description of the 20 Participating PICU

 

   DISCUSSION
TOP
 ABSTRACT
 METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
In this cohort of 13017 emergent PICU admissions, patients admitted from outside hospitals tended to have higher use of PICU-specific therapies and were more ill than patients admitted from within the same institution. However, we found that regardless of the source of admission PICU risk-adjusted mortality rates were similar.

The finding that patients admitted from outside hospitals tend to be more critically ill and require more PICU-specific therapies, although surprising in some regards, could suggest successful regionalization of pediatric critical care services in the United States today. Outlying hospitals seem to be appropriately triaging patients requiring ICU level care to centralized PICU services. Alternatively, these findings may also be the result of organizational inconsistencies within PICUs themselves, such as preferential admission to the ICU of patients from within hospitals compared with patients from outside hospitals. These differences could also be secondary to an underestimation of the severity of illness or inadequate treatment of critically ill children at remote hospitals before transfer to tertiary facilities or physiologic deterioration of patients before or during transport.

Our finding of a high acuity level of patients transferred to this sample of US PICUs is in agreement with a previous single-center study in the United Kingdom. Goh and Mok9 performed a prospective study that looked at patients transported to a tertiary PICU and found that 94% of transferred patients had either a standardized mortality risk of >1% or required ICU-dependent therapies within their first 24 hours of ICU stay. The absence of admissions from an in-house ED and the small number of in-hospital admissions from the inpatient wards in their study limited the ability to compare PICU efficiency between patients transferred from outside facilities and intrahospital admissions. We are unaware of any previous multicenter studies examining the relationship between interhospital transfer and ICU resource use in pediatrics.

The finding of significant variability at the hospital level in use of PICU-specific therapies and admission severity of illness among referred and in-house admissions was not surprising. These findings concur with a large study demonstrating significant regional variation in the hospitalization rates for all pediatric patients.12

Our results are consistent with studies in the adult ICU literature. Combes et al13 performed a retrospective review of a single-center's experience with 3416 patients admitted to the ICU, including 1005 transfer patients. Patients transferred from outside facilities experienced longer mean ICU stays (LOS), longer duration of mechanical ventilation, and greater use of ICU-dependent therapies than in-house admission patients. Similarly, Flabouris14 found an association between transfer patients and significantly higher severity of illness measures and ICU LOS. Of interest is that these studies also elicited an association between interhospital transfer and ICU mortality, a finding not apparent in our multicenter PICU population. However, a recently published study from the United States examining adult transfer patients demonstrated very similar results to our pediatric population with higher resource use and ICU LOS but similar severity-adjusted mortality rates for transfer admissions and direct admissions.15

Although our study contains data from a large multicenter, previously well-validated database and is representative of a broad cross-section of pediatric intensive care in the United States, several limitations exist to our current analysis. First, there are several variables pertaining to transfer patients who are not currently recorded in the PICUEs database. The distance from the referral hospital to the PICU, the length of transport time, and the composition of the transport team are not available for analysis in our cohort. These factors have been shown previously to correlate with the possibility for either physiologic deterioration or equipment-related mishaps,16 with specialized pediatric retrieval teams reducing the risk of patient deterioration during transport.17–19 In a study of adult ICU patients transferred from outside hospitals, travel distance was an independent predictor of overall hospital mortality.20

Another potential confounding factor in our study is the lack of information in our database about the length of time between presentation for medical care and arrival in the PICU. This raises the possibility that interventions performed before PICU admission may alter physiologic variables used in the PRISM III scoring system, limiting its ability to accurately predict mortality in this population. This "lead-time bias" could result in an underestimation of the severity of illness of transferred patients in our results, and the true difference in acuity of transported patients might be significantly greater. In an adult ICU study, the inclusion of data collected before ICU admission significantly increased severity of illness scores and estimated risk of hospital mortality.21 However, similar findings were not demonstrated in a pediatric study.22

Another potential limitation is the nature of the ICUs that constitute the PICUEs database. The PICUEs database consists of a voluntary, self-selected group of PICUs that might fundamentally differ from other PICUs across the country. Were this to be the case, it would limit the external validity of our study. However, the range of sizes, patient mixes, and geographic locations of the PICUs in the database makes it likely that these units are at least moderately similar to many ICUs across the United States to whom our findings would be generalizable.

Given the higher severity of illness of patients transferred to PICUs from outside hospitals, identifying and addressing the factors that contribute to this disparity are essential. Our results imply that patients transferred from outlying hospitals have a high level of severity of illness and ICU-level resource use, perhaps that such patients might benefit from more aggressive stabilization before transfer. The use of telemedicine technology to provide emergency and critical care consultations to remote hospitals23–27 and the availability of specialized pediatric transport services17,18 have been demonstrated to be both feasible and to improve outcomes in single-center studies. Additional multicenter studies comparing the effectiveness of such services on pre-PICU hospital resuscitation and subsequent patient outcomes would be beneficial.


   FOOTNOTES
 
Accepted Sep 21, 2007.

Address correspondence to Christopher J. Gregory, MD, University of California Davis Medical Center, 2516 Stockton Blvd, Sacramento, CA 95817. E-mail: christopher.gregory{at}ucdmc.ucdavis.edu

The authors have indicated they have no financial relationships relevant to this article to disclose.


What's Known on This Subject

Recent studies in adult critical care populations have shown an association between interhospital transfer with increased resource use and ICU length of stay but not mortality. Whether this association is present in pediatric critical care is currently unknown.

 

What This Study Adds

This is the first multicenter study looking at the effect of interhospital transfer on patient outcome and resource use in pediatric critical care. The study demonstrates a significant association between interhospital transfer and ICU resource use and some outcome variables.

 


   REFERENCES
TOP
 ABSTRACT
 METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Pearson G, Shann F, Barry P, et al. Should paediatric intensive care be centralised? Trent versus Victoria. Lancet. 1997;349 (9060):1213 –1217[CrossRef][ISI][Medline]
  2. Pollack MM, Alexander SR, Clarke N, Ruttimann UE, Tesselaar HM, Bachulis AC. Improved outcomes from tertiary center pediatric intensive care: a statewide comparison of tertiary and nontertiary care facilities. Crit Care Med. 1991;19 (2):150 –159[ISI][Medline]
  3. Ramenofsky ML, Luterman A, Quindlen E, Riddick L, Curreri PW. Maximum survival in pediatric trauma: the ideal system. J Trauma. 1984;24 (9):818 –823[ISI][Medline]
  4. Pollack MM, Cuerdon TT, Patel KM, Ruttimann UE, Getson PR, Levetown M. Impact of quality-of-care factors on pediatric intensive care unit mortality. JAMA. 1994;272 (12):941 –946[Abstract]
  5. Pollack MM, Katz RW, Ruttimann UE, Getson PR. Improving the outcome and efficiency of intensive care: the impact of an intensivist. Crit Care Med. 1988;16 (1):11 –17[ISI][Medline]
  6. Gemke RJ, Bonsel GJ. Comparative assessment of pediatric intensive care: a national multicenter study. Pediatric Intensive Care Assessment of Outcome (PICASSO) Study Group. Crit Care Med. 1995;23 (2):238 –245[CrossRef][ISI][Medline]
  7. Consensus report for regionalization of services for critically ill or injured children. Council of the Society of Critical Care Medicine. Crit Care Med. 2000;28 (1):236 –239[CrossRef][ISI][Medline]
  8. Pollack MM, Patel KM, Ruttimann UE. PRISM III: an updated Pediatric Risk of Mortality score. Crit Care Med. 1996;24 (5):743 –752[CrossRef][ISI][Medline]
  9. Goh AY, Mok Q. Centralization of paediatric intensive care: are critically ill children appropriately referred to a regional centre? Intensive Care Med. 2001;27 (4):730 –735[CrossRef][ISI][Medline]
  10. White JR, Pollack MM. Easy access to pediatric intensive care unit care: an urban legend? Crit Care Med. 2002;30 (1):256 –257[CrossRef][ISI][Medline]
  11. Kanter RK. Regional variation in child mortality at hospitals lacking a pediatric intensive care unit. Crit Care Med. 2002;30 (1):94 –99[CrossRef][ISI][Medline]
  12. Perrin JM, Homer CJ, Berwick DM, Woolf AD, Freeman JL, Wennberg JE. Variations in rates of hospitalization of children in three urban communities. N Engl J Med. 1989;320 (18):1183 –1187[Abstract]
  13. Combes A, Luyt CE, Trouillet JL, Chastre J, Gibert C. Adverse effect on a referral intensive care unit's performance of accepting patients transferred from another intensive care unit. Crit Care Med. 2005;33 (4):705 –710[CrossRef][ISI][Medline]
  14. Flabouris A. Patient referral and transportation to a regional tertiary ICU: patient demographics, severity of illness and outcome comparison with non-transported patients. Anaesth Intensive Care. 1999;27 (4):385 –390[ISI][Medline]
  15. Golestanian E, Scruggs JE, Gangnon RE, Mak RP, Wood KE. Effect of interhospital transfer on resource utilization and outcomes at a tertiary care referral center. Crit Care Med. 2007;35 (6):1470 –1476[CrossRef][ISI][Medline]
  16. Wallen E, Venkataraman ST, Grosso MJ, Kiene K, Orr RA. Intrahospital transport of critically ill pediatric patients. Crit Care Med. 1995;23 (9):1588 –1595[CrossRef][ISI][Medline]
  17. Britto J, Nadel S, Maconochie I, Levin M, Habibi P. Morbidity and severity of illness during interhospital transfer: impact of a specialised paediatric retrieval team. BMJ. 1995;311 (7009):836 –839[Abstract/Free Full Text]
  18. Edge WE, Kanter RK, Weigle CG, Walsh RF. Reduction of morbidity in interhospital transport by specialized pediatric staff. Crit Care Med. 1994;22 (7):1186 –1191[ISI][Medline]
  19. Vos GD, Nissen AC, Nieman FH, et al. Comparison of interhospital pediatric intensive care transport accompanied by a referring specialist or a specialist retrieval team. Intensive Care Med. 2004;30 (2):302 –308[CrossRef][ISI][Medline]
  20. Durairaj L, Will JG, Torner JC, Doebbeling BN. Prognostic factors for mortality following interhospital transfers to the medical intensive care unit of a tertiary referral center. Crit Care Med. 2003;31 (7):1981 –1986[CrossRef][ISI][Medline]
  21. Tunnell RD, Millar BW, Smith GB. The effect of lead time bias on severity of illness scoring, mortality prediction and standardised mortality ratio in intensive care: a pilot study. Anaesthesia. 1998;53 (11):1045 –1053[CrossRef][ISI][Medline]
  22. Orr RA, Venkataraman ST, Cinoman MI, Hogue BL, Singleton CA, McCloskey KA. Pretransport Pediatric Risk of Mortality (PRISM) score underestimates the requirement for intensive care or major interventions during interhospital transport. Crit Care Med. 1994;22 (1):101 –107[ISI][Medline]
  23. Marcin JP, Nesbitt TS, Kallas HJ, Struve SN, Traugott CA, Dimand RJ. Use of telemedicine to provide pediatric critical care inpatient consultations to underserved rural Northern California. J Pediatr. 2004;144 (3):375 –380[CrossRef][ISI][Medline]
  24. Rogers FB, Ricci M, Caputo M, et al. The use of telemedicine for real-time video consultation between trauma center and community hospital in a rural setting improves early trauma care: preliminary results. J Trauma. 2001;51 (6):1037 –1041[ISI][Medline]
  25. Rosenfeld BA, Dorman T, Breslow MJ, et al. Intensive care unit telemedicine: alternate paradigm for providing continuous intensivist care. Crit Care Med. 2000;28 (12):3925 –3931[CrossRef][ISI][Medline]
  26. Ricci MA, Caputo M, Amour J, et al. Telemedicine reduces discrepancies in rural trauma care. Telemed J E Health. 2003;9 (1):3 –11[CrossRef][ISI][Medline]
  27. Kofos D, Pitetti R, Orr R, Thompson A. Telemedicine in pediatric transport: a feasibility study. Pediatrics. 1998;102 (5). Available at: www.pediatrics.org/cgi/content/full/102/5/e58
PEDIATRICS (ISSN 1098-4275). ©2008 by the American Academy of Pediatrics




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow P3Rs: Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when P3Rs are posted
Right arrow Alert me if a correction is posted
Services
Right arrow E-mail this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My File Cabinet
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via CrossRef
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Gregory, C. J.
Right arrow Articles by Marcin, J. P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Gregory, C. J.
Right arrow Articles by Marcin, J. P.
Related Collections
Right arrow Emergency Medicine