0
Original Research: CRITICAL CARE |

Impact of Nonphysician Staffing on Outcomes in a Medical ICU FREE TO VIEW

Hayley B. Gershengorn, MD; Hannah Wunsch, MD; Romina Wahab, MD; David Leaf, MD; Daniel Brodie, MD, FCCP; Guohua Li, MD, DrPH; Phillip Factor, DO, FCCP
Author and Funding Information

From the Division of Pulmonary, Critical Care, and Sleep Medicine (Drs Gershengorn and Factor), Beth Israel Medical Center; and the Department of Anesthesia (Drs Wunsch and Li), the Department of Medicine (Drs Wahab and Leaf), and the Division of Pulmonary, Allergy, and Critical Care (Dr Brodie), New York Presbyterian Hospital-Columbia, New York, NY.

Correspondence to: Hayley B. Gershengorn, MD, Section of Pulmonary, Critical Care, and Sleep Medicine, Beth Israel Medical Center, 1st Ave at 16th St, New York, NY 10003; e-mail: hgershengorn@chpnet.org


This work has been presented in poster form (Gershengorn HB, Wunsch H, Wahab R, et al. Am J Respir Crit Care Med. 2010:A2409).

Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (http://www.chestpubs.org/site/misc/reprints.xhtml).


© 2011 American College of Chest Physicians


Chest. 2011;139(6):1347-1353. doi:10.1378/chest.10-2648
Text Size: A A A
Published online

Background:  As the number of ICU beds and demand for intensivists increase, alternative solutions are needed to provide coverage for critically ill patients. The impact of different staffing models on the outcomes of patients in the medical ICU (MICU) remains unknown. In our study, we compare outcomes of nonphysician provider-based teams to those of medical house staff-based teams in the MICU.

Methods:  We conducted a retrospective review of 590 daytime (7:00 am-7:00 pm) admissions to two MICUs at one hospital. In one MICU staffed by nurse practitioners and physician assistants (MICU-NP/PA) there were nonphysicians (nurse practitioners and physicians assistants) during the day (7:00 am-7:00 pm) with attending physician coverage overnight. In the other MICU, there were medicine residents (MICU-RES) (24 h/d). The outcomes investigated were hospital mortality, length of stay (LOS) (ICU, hospital), and posthospital discharge destination.

Results:  Three hundred two patients were admitted to the MICU-NP/PA and 288 to the MICU-RES. Mortality probability model III (MPM0-III) predicted mortality was similar (P = .14). There was no significant difference in hospital mortality (32.1% for MICU-NP/PA vs 32.3% for MICU-RES, P = .96), MICU LOS (4.22 ± 2.51 days for MICU-NP/PA vs 4.44 ± 3.10 days for MICU-RES, P = .59), or hospital LOS (14.01 ± 2.92 days for MICU-NP/PA vs 13.74 ± 2.94 days for MICU-RES, P = .86). Discharge to a skilled care facility (vs home) was similar (37.1% for MICU-NP/PA vs 32.5% for MICU-RES, P = .34). After multivariate adjustment, MICU staffing type was not associated with hospital mortality (P = .26), MICU LOS (P = .29), hospital LOS (P = .19), or posthospital discharge destination (P = .90).

Conclusions:  Staffing models including daytime use of nonphysician providers appear to be a safe and effective alternative to the traditional house staff-based team in a high-acuity, adult ICU.

Figures in this Article

Worldwide trends, professional societies such as the Society of Critical Care Medicine, and regulatory agencies such as the Leapfrog Group increasingly expect ICUs to have high-intensity coverage (board-certified intensivists providing patient care for extended periods of the day).1,2 There are, however, an insufficient number of such intensivists.3 The expectation that they can provide coverage for all critically ill patients is not currently feasible.

Alternative staffing solutions are required. Many hospitals deliver critical care using physicians-in-training. Increasingly, however, patient care by physicians-in-training is being curtailed because of work hour limitations and prioritization of education over service delivery.4,5 Some hospitals use hospitalists to provide critical care services.6 However, like intensivists, hospitalists are in short supply and neither meet the recommendations of regulatory agencies nor provide the specialized care needed by many ICU patients. The combination of increasing demands for critical care coverage, reduced availability of physicians-in-training, and increasingly insufficient numbers of board-certified intensivists has prompted hospitals to consider employing nonphysician providers (nurse practitioners [NPs] and physician assistants [PAs]) under the supervision of attending physicians to augment or replace doctors in the ICU.7

This use of nonphysician care providers to address physician staffing shortages is well established on the general hospital ward.813 Within the ICU, nonphysician providers are an established presence in the care of neonates and children. As early as 1994, a survey revealed that close to one-half of all hospitals used NPs/PAs in the care of children, with the vast majority of these caregivers employed in ICUs.14 Moreover, NPs/PAs have been found to positively impact care in pediatric/neonatal ICUs.15

Information on the impact of NPs and PAs in the care of adult patients with critical illness is less robust.16 A single study assessed the impact of nonphysician providers on the acutely ill patient in the medical ICU (MICU).17 MICU lengths of stay and mortality comparable to historical data were observed, but there were no concurrent control groups with whom to compare outcome measures.

In the winter of 2005, the Columbia Presbyterian Medical Center expanded its medical critical care services by opening a 12-bed MICU staffed by NPs and PAs under the direction of board-certified intensivists and critical care fellows. To evaluate the effectiveness of critical care delivered in this manner, we compared outcomes in this new nonphysician provider-staffed MICU to an otherwise similar MICU at the same institution staffed by medical residents. The data presented here show that commonly measured outcomes were similar in the two units. These findings are the first to show equivalency of care provided by a team structured around nonphysician providers in a high-intensity ICU environment and, thus, lend substantive support to the use of NPs and PAs in the care of acutely critically ill patients in a tertiary care setting.

Study Design

This is a retrospective study of admissions to two 12-bed MICUs at Columbia Presbyterian Medical Center. One is staffed by NPs and PAs (MICU-NP/PA) during the day (7:00 am-7:00 pm), with in-unit board-certified intensivist coverage overnight. The other is staffed by first- and second-year medicine residents (MICU-RES) at all times with off-site fellow and board-certified intensivist back-up at night (7:00 pm-7:00 am). Both units have an on-site fellow and intensivist during the day (7:00 am-7:00 pm).

Ten board-certified acute care NPs and three to four board-certified PAs staff the MICU-NP/PA. Each works 13 12-h shifts per month. Most are hired immediately following graduation or with minimal previous critical care experience. All undergo a 2- to 4-week orientation with experienced colleagues providing bedside mentorship. Medicine house staff work in the MICU-RES. They spend one 4-week block in each of their first and second residency years in the MICU-RES. From 7:00 am to 7:00 pm there are two to three first-year and two to three second-year residents in the unit; overnight there is one first-year and one second-year resident. House staff take overnight call every fourth night. The nighttime intensivists work on a voluntary basis, averaging two to three shifts per month. Pulmonary and Critical Care Medicine fellows (rotating monthly), board-certified intensivists (rotating weekly), nurses, and respiratory therapists work interchangeably in either unit; in both units there are the same caregiver-to-patient ratios (1:2 for nurses and 1:10-20 for respiratory therapists). Daily rounds are open to all caregivers but are attended primarily by physicians. Both units are overseen by the same nurse manager and pulmonary and critical care board-certified MICU director.

One other difference between units is that the MICU-RES accepts new admissions 24 h/d, whereas the MICU-NP/PA only admits patients 7:00 am-7:00 pm. Patients are transferred from MICU-RES to MICU-NP/PA overnight as required to allow for new admissions. Triage during daytime hours is based largely on bed availability; no strict guidelines exist for placement in either unit.

Patients and Patient Data

Patients admitted to both MICUs from January 1 to December 31, 2008 were included. This time frame was selected to account for variation in house staff skills over the year. We excluded patients if they were admitted overnight (7:00 pm-7:00 am) since they could only be admitted to the MICU-RES. If patients were admitted to the MICU-RES and transferred to the MICU-NP/PA, they were categorized as MICU-RES patients (intention to treat); however, their MICU length of stay included time spent in either unit. Only the initial MICU admission of a hospitalization was included because the primary outcome, hospital mortality, could only occur for each patient once per hospitalization (not on each MICU admission).

We obtained data (age, race, gender, ICU admitting diagnosis, code status, location prior to MICU, hospital and MICU disposition, and lengths of stay [LOS]) from review of electronic medical records. One author (H. G.) calculated mortality probability model III (MPM0-III) scores retrospectively using computerized notes.18 The MPM0-III score was chosen because it could be accurately calculated retrospectively from available medical records.

Statistical Analysis

We summarized patient characteristics in each MICU using percentages or means ± SDs and compared them using Student t tests or χ2 tests as appropriate. We investigated the impact of being in the MICU-NP/PA vs the MICU-RES using multivariate logistic and linear regression. Our primary outcome was in-hospital mortality. Secondary outcomes included MICU LOS, hospital LOS, and functional status (if alive) on hospital discharge (by discharge home vs to a skilled care facility). LOS variables were logarithmically transformed to better conform to a normal distribution. Univariate analyses were conducted for all aforementioned patient characteristics. Multivariate models were built using backward stepwise elimination, initially including all variables for which P < .2 in univariate analysis. Only reported variables were included in the multivariate analysis. Because MPM0-III is calculated using several model variables (age, code status, and certain acute diagnoses), we reran the analyses adjusting for components of the MPM0-III score separately to confirm we had not overfit our model. Additionally, we recreated our model excluding patients transferred between units to confirm that their inclusion did not significantly alter the results. Given our sample size and observed hospital mortality, our study can detect a 10% difference in mortality with 80% power.

Database management and statistical analyses were performed using Excel (Microsoft; Redmond, Washington) and Stata 10.0 (StataCorp LP; College Station, Texas). The study was reviewed and approved by the institutional review board of Columbia Presbyterian Medical Center (IRB-AAAE0445). Data were collected retrospectively after institutional review board approval was obtained.

MICU Case Mix

There were a total of 1,002 admissions to the two MICUs in 2008. Of these, 590 (58.9%) were admitted during the day (7:00 am-7:00 pm) (Fig 1). Of those patients admitted during the day, 302 (51.2%) were admitted to the MICU-NP/PA and 288 (48.8%) to the MICU-RES. Forty-one (14.2%) of the patients initially admitted during the day to the MICU-RES were transferred to the MICU-NP/PA during their MICU stay. Patients admitted to the MICU-NP/PA were older (64.7 ± 17.0 years vs 57.8 ± 18.1 years in the MICU-RES, P < .01) and more likely to have a do-not-resuscitate (DNR) order in place on admission to the ICU (6.3% vs 1.4%, P < .01) (Table 1). There were no significant differences in severity of illness, as assessed by MPM0-III prediction of in-hospital mortality, gender, race, likelihood of having a DNR order placed during the ICU stay, or admitting diagnosis.

Figure Jump LinkFigure 1. Patient admissions to two MICUs with different staffing patterns from January 1, 2008, to December 31, 2008. *Including only first admissions to the MICU for a given hospital stay. **These groups define MICU-NP/PA and MICU-RES for primary analyses. MICU = medical ICU; MICU-NP/PA = medical ICU staffed by nurse practitioners and physician assistants; MICU-RES = medical ICU staffed by medical residents.Grahic Jump Location
Table Graphic Jump Location
Table 1 —Baseline Characteristics of Patients Cared for in Two MICUs With Different Staffing Patterns

Data are presented as No. (%) or mean ± SD. DNR = do not resuscitate; MICU = medical ICU; MICU-NP/PA = medical ICU staffed by nurse practitioners and physician assistants; MICU-RES = medical ICU staffed by medical residents; MPM0-III = mortality probability model III.

a 

Of χ2 or t test analysis for comparisons between MICU-RES and MICU-NP/PA.

b 

ED = patients who proceed directly from the ED to the MICU; Hospital ward = those in a regular (non-ICU) hospital bed at our institution; Other = other ICUs, other hospitals, direct admissions to the MICU, and so forth.

c 

Predictions derived from MPM0-III.

d 

Based on International Classification of Diseases, Ninth Revision coding. Bleed = bleeding from any source; GI = nonbleeding/noninfectious problem arising from an organ within the GI system; Infection = infection originating in any organ; Other = diagnoses not fitting within another category; Respiratory = nonbleeding/noninfectious problem arising from the lungs.

Patient Outcomes

There was no significant difference in our primary outcome of unadjusted hospital mortality in the two MICUs (32.1% in MICU-NP/PA vs 32.3% in MICU-RES, P = .96) (Table 2). ICU LOS and hospital LOS were also similar in the two MICUs (ICU LOS, 4.22 ± 2.51 days for MICU-NP/PA vs 4.44 ± 3.10 days for MICU-RES, P = .59; and hospital LOS 14.01 ± 2.92 days for MICU-NP/PA vs 13.74 ± 2.94 days for MICU-RES, P = .86). ICU mortality did not differ between the two units (19.5% for MICU-NP/PA vs 20.8% for MICU-RES, P = .70). Patients who survived to hospital discharge were just as likely to be discharged to a skilled care facility after the hospital stay (37.1% in MICU-NP/PA vs 32.5% in MICU-RES, P = .32).

Table Graphic Jump Location
Table 2 —Outcomes for Patients Cared for in MICUs With Different Staffing Patterns

ORs and coefficients refer to impact of being in the MICU-NP/PA on each outcome (reference group = MICU-RES). LOS = length of stay; SNF = skilled nursing facility. See Table 1 legend for expansion of other abbreviations.

a 

Adjustments made for age, gender, race, resuscitation status, location prior to MICU, MPM0-III mortality probability, and admitting diagnosis.

b 

For all patients.

c 

For only hospital survivors; SNF refers to any facility with skilled nursing for the short- or long-term (eg, nursing homes, rehabilitation facilities, other acute care hospitals); reference group = patients discharged directly home.

d 

Coefficients correspond to odds of increasing MICU LOS by 1 day of 0.95 (0.80, 1.13) (unadjusted) and 0.93 (0.83, 1.06) (adjusted).

e 

Coefficients correspond to odds of increasing hospital LOS by 1 day of 1.02 (0.85, 1.21) (unadjusted) and 1.07 (0.95, 1.21) (adjusted).

Multivariate modeling revealed no statistically significant difference in hospital mortality for patients cared for in MICU-NP/PA vs MICU-RES (adjusted OR, 0.75; 95% CI, 0.46-1.23; P = .26) (Table 2). After multivariate adjustment, there were also no statistically significant differences between MICU-NP/PA and MICU-RES in any of the secondary outcomes (Table 2); being in MICU-NP/PA did not impact MICU LOS (P = .29), hospital LOS (P = .19), or posthospital discharge destination for survivors (P = .90).

Variables that were predictive of significantly increased risk of in-hospital mortality were having a DNR order in place on admission to the ICU (OR, 6.94; P < .01), having a DNR order placed during the ICU stay (OR, 25.26; P < .01), admission to the ICU from a location other than the ED (hospital ward: OR, 2.34; P < .01; other: OR, 3.09; P < .01), and higher severity of illness (ref MPM0-III < 25%; MPM0-III 25%-49%: OR, 3.48; P < .01; MPM0-III 50% to 74%: OR, 2.21; P = .11; MPM0-III ≥ 75%: OR, 11.96; P = .03) (Table 3). Patients with a primary diagnosis of bleeding were significantly less likely to die during the hospitalization compared with patients with other diagnoses (OR, 0.06; P = .01).

Table Graphic Jump Location
Table 3 —Multivariate Modeling of Risk Factors for Hospital Mortality

Variables not included in this analysis were removed following model construction using backward stepwise elimination, initially including all variables for which P < .2 in univariate analysis. MPM0-III = mortality probability model III. See Table 1 for expansion of other abbreviations.

a 

P value for the individual coefficients in the multivariate regression analysis; the P value for the complete multivariate regression model < .01.

b 

Refers to odds associated with being in MICU-NP/PA (reference group = MICU-RES).

c 

See footnote b in Table 1 for a description of the locations.

d 

Based on International Classification of Diseases, Ninth Revision coding. See footnote d in Table 1 for an explanation of the diagnosis terms.

Exclusion of Transferred Patients

Sensitivity analysis excluding patients transferred between units showed no statistical differences between groups (data not shown). Multivariate analyses excluding transfers also revealed no association between being in the MICU-NP/PA and any evaluated outcome.

Optimum staffing models for the delivery of critical care are unknown, while, simultaneously, demand for critical care services is growing. Our findings suggest that an acute care NP/PA-based team can provide critical care to severely ill patients in an MICU in a large urban academic environment with no significant difference in clinically relevant measurable outcomes when compared with a traditional, house staff-based team. Models using nonphysician providers, therefore, may be one solution to the increased demand for critical care practitioners.

High-intensity ICU staffing by board-certified intensivists is associated with improved clinical outcomes19 and is the standard set forth by the Society of Critical Care Medicine and the Leapfrog Group.1,2 Such a model demands, at a minimum, mandatory critical care consultation for all patients but, in many instances, implies the presence of a board-certified intensivist in the ICU at all times of the day. In the MICU-NP/PA, attending physicians are the primary critical care providers overnight. In an environment where ICU physician supply is constrained, however, it is cost and manpower inefficient to have an attending intensivist provide first-line care. As such, other care models must be considered.

Nonphysician providers have been employed in ICUs7 to fill the staffing gaps due to expansions in the number of ICU beds,20 shortages of intensivists,21 and reductions in house staff.16 Pediatric and neonatal ICUs were the pioneers in this arena with studies from the mid-1990s showing similar morbidity and mortality for patients cared for by nonphysician providers to those cared for by pediatric residents.22,23 Two previous studies examined the impact of staffing with NPs and PAs on the adult medical critically ill patient. Hoffman et al24,25 looked at long-stay patients in the acute and subacute MICUs at the University of Pittsburgh cared for by either an “attending plus fellow” or an “attending plus NP” team and found no difference in ventilator weaning rates, LOS, or death. In the subacute MICU, NPs spent more time speaking with families, whereas house staff attended conferences more frequently.26 Dubaybo et al17 conducted a study (using historical control groups) that directly compared staffing with PAs to staffing with residents and found no difference in MICU LOS or mortality rates for patients cared for by PAs. In this study, however, the patients cared for by house staff had a higher severity of illness, making comparisons of care among the groups impossible. Furthermore, the in-MICU mortality rates (41%) in this analysis were higher than in current ICU studies,18,27,28 raising questions about the generalizability of these findings. The unique structure of the medical ICUs at our institution allows for a comparison of outcomes that does not require historical control groups. Moreover, the similar patient mix and severity of illness in the two units limits potential biases often found in retrospective cohort studies. These aspects of the study design strengthen our findings that a critical care team using NPs and PAs provides comparable care to a house staff-based team.

Our study has a number of limitations. First, it is possible that the presence of an on-site intensivist at night (primarily responsible for MICU-NP/PA) drives the clinical success of the MICU-NP/PA. Although studies suggest that better hospital survival is seen with high-intensity ICU staffing,19 to our knowledge, there are only two single-center studies in the literature specifically addressing the clinical impact of a 24-h in-house intensivist presence.29 In one, although hospital LOS and in-ICU complications were found to be decreased with 24-h on-site coverage, there was no difference in mortality (hospital or ICU)30; in the other, however, hospital mortality improved.31 Further research is clearly needed to explore the potential importance of an on-site intensivist 24 h/d vs one on-site during the daytime only.

Second, our results could be confounded by the different admissions policies in the two MICUs. In particular, only one unit (MICU-RES) admitted patients overnight, mandating transfer of patients to the other unit (MICU-NP/PA) during this time to create room for new admissions. A number of studies have investigated the impact of being admitted to an ICU at night with mixed results regarding whether there is harm, benefit, or no effect associated with admission time.3235 Only one study looked specifically at MICU patients and found no association between timing of ICU admission and likelihood of hospital mortality.36 Consequently, our exclusion of patients admitted overnight likely had little to no impact on our results. To address the impact of patients transferred between units, we performed a sensitivity analysis excluding them, in which we found, again, no significant difference in clinical outcomes. However, we cannot be sure that the presence of these “crossover” patients does not affect our results in some unquantifiable way.

Third, our study is a retrospective review and not a randomized controlled trial. Thus, it is possible that there are inherent differences between patient groups that may impact our results. Among reviewed patient characteristics, however, we found that patients in the MICU-NP/PA tended to be older, more likely to come from the hospital floor than the ED (a factor known to be associated with higher mortality27), and more likely to have a DNR order in place at the time of ICU admission. Each of these differences would suggest that the population cared for in the MICU-NP/PA was less likely to do well than those patients cared for in the MICU-RES and, hence, would bias against our findings of equivalent care.

Fourth, our study is of patients in a single large academic medical center. In particular, our study groups had higher observed hospital mortality than predicted by MPM0-III. This finding brings up the question of whether our results are transferrable to units with different observed to expected mortality ratios at baseline.

Finally, our two MICUs have only existed in their current forms for a short time. Consequently, we were limited in the number of patients we could analyze. As such, we only had power to detect relatively large differences in outcomes between the two units. Moreover, nonphysician providers were employed only during the daytime. Other studies are clearly needed to confirm our findings and to explore alternative staffing options, including extending nonphysician provider coverage to nighttime hours.

Nonphysician intensive care providers are often relegated to care for less ill patients or are used to supplement care provided by house staff. The data herein are strong evidence that teams centered around nonphysician intensivists, when accompanied by appropriate physician leadership, can provide critical care that is equivalent to that delivered by teams built around house staff in a high-intensity, tertiary care setting with very ill patients. Worsening intensivist staffing shortages in a setting of growing demand for intensive care in the United States mandates the development of new models for the care of the critically ill that better use available human resources. Our data support the use of NPs and PAs in these new models of care.

Author contributions: Dr Gershengorn had full access to all data and can vouch for the integrity of all analyses.

Dr Gershengorn: contributed to data collection, data analysis, and manuscript preparation.

Dr Wunsch: contributed to data analysis and manuscript drafting.

Dr Wahab: contributed to data collection and manuscript revision.

Dr Leaf: contributed to data collection and manuscript revision.

Dr Brodie: contributed to data collection and manuscript revision.

Dr Li: contributed to data analysis and manuscript drafting.

Dr Factor: contributed to data analysis and manuscript drafting.

Financial/nonfinancial disclosures: The authors have reported to CHEST that no potential conflicts of interest exist with any companies/organizations whose products or services may be discussed in this article.

Other contributions: This work was performed at the New York Presbyterian Hospital—Columbia University Medical Center.

DNR

do not resuscitate

LOS

length of stay

MICU

medical ICU

MICU-NP/PA

medical ICU staffed by nurse practitioners and physician assistants

MICU-RES

medical ICU staffed by medical residents

MPM0-III

mortality probability model III

NP

nurse practitioner

PA

physician assistant

Haupt MT, Bekes CE, Brilli RJ, et al; Task Force of the American College of Critical Care Medicine, Society of Critical Care Medicine Task Force of the American College of Critical Care Medicine, Society of Critical Care Medicine Guidelines on critical care services and personnel: Recommendations based on a system of categorization of three levels of care. Crit Care Med. 2003;3111:2677-2683. [CrossRef] [PubMed]
 
ICU physician staffing factsheetICU physician staffing factsheet The LeapFrog Group Web site.http://www.leapfroggroup.org/media/file/FactSheet_IPS.pdf. Accessed December 17, 2009.
 
Angus DC, Kelley MA, Schmitz RJ, White A, Popovich J Jr. Committee on Manpower for Pulmonary and Critical Care Societies (COMPACCS) Committee on Manpower for Pulmonary and Critical Care Societies (COMPACCS) Caring for the critically ill patient. Current and projected workforce requirements for care of the critically ill and patients with pulmonary disease: can we meet the requirements of an aging population? JAMA. 2000;28421:2762-2770. [CrossRef] [PubMed]
 
The ACGME’s approach to limit resident duty hours 2007-08: a summary of achievements for the fifth year under the common requirements, 2007-2008. Accreditation Council for Graduate Medical Education Web site.http://www.acgme.org/acWebsite/dutyHours/dh_achievesum0708.pdf. Accessed November 18, 2009.
 
Iglehart JK. Revisiting duty-hour limits—IOM recommendations for patient safety and resident education. N Engl J Med. 2008;35925:2633-2635. [CrossRef] [PubMed]
 
Heisler M. Hospitalists and intensivists: partners in caring for the critically ill—the time has come. J Hosp Med. 2010;51:1-3. [CrossRef] [PubMed]
 
D’Agostino R, Halpern NA. Acute care nurse practitioners in oncologic critical care: the memorial Sloan-Kettering cancer center experience. Crit Care Clin. 2010;261:207-217. [CrossRef] [PubMed]
 
Resnick AS, Todd BA, Mullen JL, Morris JB. How do surgical residents and non-physician practitioners play together in the sandbox? Curr Surg. 2006;632:155-164. [CrossRef] [PubMed]
 
Stahlfeld KR, Robinson JM, Burton EC. What do physician extenders in a general surgery residency really do? J Surg Educ. 2008;655:354-358. [CrossRef] [PubMed]
 
Christmas AB, Reynolds J, Hodges S, et al. Physician extenders impact trauma systems. J Trauma. 2005;585:917-920. [CrossRef] [PubMed]
 
Oswanski MF, Sharma OP, Raj SS. Comparative review of use of physician assistants in a level I trauma center. Am Surg. 2004;703:272-279. [PubMed]
 
Victorino GP, Organ CH Jr. Physician assistant influence on surgery residents. Arch Surg. 2003;1389:971-975. [CrossRef] [PubMed]
 
Lundberg S, Wali S, Thomas P, Cope D. Attaining resident duty hours compliance: the acute care nurse practitioners program at Olive View-UCLA Medical Center. Acad Med. 2006;8112:1021-1025. [CrossRef] [PubMed]
 
DeNicola L, Kleid D, Brink L, et al. Use of pediatric physician extenders in pediatric and neonatal intensive care units. Crit Care Med. 1994;2211:1856-1864. [PubMed]
 
Cramer CL, Orlowski JP, DeNicola LK. Pediatric intensivist extenders in the pediatric ICU. Pediatr Clin North Am. 2008;553:687-708. [CrossRef] [PubMed]
 
Kleinpell RM, Ely EW, Grabenkort R. Nurse practitioners and physician assistants in the intensive care unit: an evidence-based review. Crit Care Med. 2008;3610:2888-2897. [CrossRef] [PubMed]
 
Dubaybo BA, Samson MK, Carlson RW. The role of physician-assistants in critical care units. Chest. 1991;991:89-91. [CrossRef] [PubMed]
 
Higgins TL, Teres D, Copes WS, Nathanson BH, Stark M, Kramer AA. Assessing contemporary intensive care unit outcome: an updated Mortality Probability Admission Model (MPM0-III). Crit Care Med. 2007;353:827-835. [CrossRef] [PubMed]
 
Pronovost PJ, Angus DC, Dorman T, Robinson KA, Dremsizov TT, Young TL. Physician staffing patterns and clinical outcomes in critically ill patients: a systematic review. JAMA. 2002;28817:2151-2162. [CrossRef] [PubMed]
 
Halpern NA, Pastores SM. Critical care medicine in the United States 2000-2005: an analysis of bed numbers, occupancy rates, payer mix, and costs. Crit Care Med. 2010;381:65-71. [CrossRef] [PubMed]
 
Angus DC, Shorr AF, White A, Dremsizov TT, Schmitz RJ, Kelley MA. Committee on Manpower for Pulmonary and Critical Care Societies (COMPACCS) Committee on Manpower for Pulmonary and Critical Care Societies (COMPACCS) Critical care delivery in the United States: distribution of services and compliance with Leapfrog recommendations. Crit Care Med. 2006;344:1016-1024. [CrossRef] [PubMed]
 
Carzoli RP, Martinez-Cruz M, Cuevas LL, Murphy S, Chiu T. Comparison of neonatal nurse practitioners, physician assistants, and residents in the neonatal intensive care unit. Arch Pediatr Adolesc Med. 1994;14812:1271-1276. [CrossRef] [PubMed]
 
Mitchell-DiCenso A, Guyatt G, Marrin M, et al. A controlled trial of nurse practitioners in neonatal intensive care. Pediatrics. 1996;986 pt 1:1143-1148. [PubMed]
 
Hoffman L, Tasota F, Zullo T, Scharfenberg C, Donahoe MP. Outcomes of care managed by an acute care nurse practitioner/attending physician team in a subacute medical intensive care unit. Am J Crit Care. 2005;142:121-130. [PubMed]
 
Hoffman LA, Miller TH, Zullo TG, Donahoe MP. Comparison of 2 models for managing tracheotomized patients in a subacute medical intensive care unit. Respir Care. 2006;5111:1230-1236. [PubMed]
 
Hoffman LA, Tasota FJ, Scharfenberg C, Zullo TG, Donahoe MP. Management of patients in the intensive care unit: comparison via work sampling analysis of an acute care nurse practitioner and physicians in training. Am J Crit Care. 2003;125:436-443. [PubMed]
 
Zimmerman JE, Kramer AA, McNair DS, Malila FM. Acute Physiology and Chronic Health Evaluation (APACHE) IV: hospital mortality assessment for today’s critically ill patients. Crit Care Med. 2006;345:1297-1310. [CrossRef] [PubMed]
 
Moran JL, Bristow P, Solomon PJ, George C, Hart GK. Australian and New Zealand Intensive Care Society Database Management Committee (ADMC) Australian and New Zealand Intensive Care Society Database Management Committee (ADMC) Mortality and length-of-stay outcomes, 1993-2003, in the binational Australian and New Zealand intensive care adult patient database. Crit Care Med. 2008;361:46-61. [CrossRef] [PubMed]
 
Kahn JM, Hall JB. More doctors to the rescue in the intensive care unit: a cautionary note. Am J Respir Crit Care Med. 2010;18111:1160-1161. [CrossRef] [PubMed]
 
Gajic O, Afessa B, Hanson AC, et al. Effect of 24-hour mandatory versus on-demand critical care specialist presence on quality of care and family and provider satisfaction in the intensive care unit of a teaching hospital. Crit Care Med. 2008;361:36-44. [CrossRef] [PubMed]
 
Blunt MC, Burchett KR. Out-of-hours consultant cover and case-mix-adjusted mortality in intensive care. Lancet. 2000;3569231:735-736. [CrossRef] [PubMed]
 
Arabi Y, Alshimemeri A, Taher S. Weekend and weeknight admissions have the same outcome of weekday admissions to an intensive care unit with onsite intensivist coverage. Crit Care Med. 2006;343:605-611. [CrossRef] [PubMed]
 
Wunsch H, Mapstone J, Brady T, Hanks R, Rowan K. Hospital mortality associated with day and time of admission to intensive care units. Intensive Care Med. 2004;305:895-901. [CrossRef] [PubMed]
 
Luyt CE, Combes A, Aegerter P, et al. Mortality among patients admitted to intensive care units during weekday day shifts compared with “off” hours. Crit Care Med. 2007;351:3-11. [CrossRef] [PubMed]
 
Laupland KB, Shahpori R, Kirkpatrick AW, Stelfox HT. Hospital mortality among adults admitted to and discharged from intensive care on weekends and evenings. J Crit Care. 2008;233:317-324. [CrossRef] [PubMed]
 
Morales IJ, Peters SG, Afessa B. Hospital mortality rate and length of stay in patients admitted at night to the intensive care unit. Crit Care Med. 2003;313:858-863. [CrossRef] [PubMed]
 

Figures

Figure Jump LinkFigure 1. Patient admissions to two MICUs with different staffing patterns from January 1, 2008, to December 31, 2008. *Including only first admissions to the MICU for a given hospital stay. **These groups define MICU-NP/PA and MICU-RES for primary analyses. MICU = medical ICU; MICU-NP/PA = medical ICU staffed by nurse practitioners and physician assistants; MICU-RES = medical ICU staffed by medical residents.Grahic Jump Location

Tables

Table Graphic Jump Location
Table 1 —Baseline Characteristics of Patients Cared for in Two MICUs With Different Staffing Patterns

Data are presented as No. (%) or mean ± SD. DNR = do not resuscitate; MICU = medical ICU; MICU-NP/PA = medical ICU staffed by nurse practitioners and physician assistants; MICU-RES = medical ICU staffed by medical residents; MPM0-III = mortality probability model III.

a 

Of χ2 or t test analysis for comparisons between MICU-RES and MICU-NP/PA.

b 

ED = patients who proceed directly from the ED to the MICU; Hospital ward = those in a regular (non-ICU) hospital bed at our institution; Other = other ICUs, other hospitals, direct admissions to the MICU, and so forth.

c 

Predictions derived from MPM0-III.

d 

Based on International Classification of Diseases, Ninth Revision coding. Bleed = bleeding from any source; GI = nonbleeding/noninfectious problem arising from an organ within the GI system; Infection = infection originating in any organ; Other = diagnoses not fitting within another category; Respiratory = nonbleeding/noninfectious problem arising from the lungs.

Table Graphic Jump Location
Table 2 —Outcomes for Patients Cared for in MICUs With Different Staffing Patterns

ORs and coefficients refer to impact of being in the MICU-NP/PA on each outcome (reference group = MICU-RES). LOS = length of stay; SNF = skilled nursing facility. See Table 1 legend for expansion of other abbreviations.

a 

Adjustments made for age, gender, race, resuscitation status, location prior to MICU, MPM0-III mortality probability, and admitting diagnosis.

b 

For all patients.

c 

For only hospital survivors; SNF refers to any facility with skilled nursing for the short- or long-term (eg, nursing homes, rehabilitation facilities, other acute care hospitals); reference group = patients discharged directly home.

d 

Coefficients correspond to odds of increasing MICU LOS by 1 day of 0.95 (0.80, 1.13) (unadjusted) and 0.93 (0.83, 1.06) (adjusted).

e 

Coefficients correspond to odds of increasing hospital LOS by 1 day of 1.02 (0.85, 1.21) (unadjusted) and 1.07 (0.95, 1.21) (adjusted).

Table Graphic Jump Location
Table 3 —Multivariate Modeling of Risk Factors for Hospital Mortality

Variables not included in this analysis were removed following model construction using backward stepwise elimination, initially including all variables for which P < .2 in univariate analysis. MPM0-III = mortality probability model III. See Table 1 for expansion of other abbreviations.

a 

P value for the individual coefficients in the multivariate regression analysis; the P value for the complete multivariate regression model < .01.

b 

Refers to odds associated with being in MICU-NP/PA (reference group = MICU-RES).

c 

See footnote b in Table 1 for a description of the locations.

d 

Based on International Classification of Diseases, Ninth Revision coding. See footnote d in Table 1 for an explanation of the diagnosis terms.

References

Haupt MT, Bekes CE, Brilli RJ, et al; Task Force of the American College of Critical Care Medicine, Society of Critical Care Medicine Task Force of the American College of Critical Care Medicine, Society of Critical Care Medicine Guidelines on critical care services and personnel: Recommendations based on a system of categorization of three levels of care. Crit Care Med. 2003;3111:2677-2683. [CrossRef] [PubMed]
 
ICU physician staffing factsheetICU physician staffing factsheet The LeapFrog Group Web site.http://www.leapfroggroup.org/media/file/FactSheet_IPS.pdf. Accessed December 17, 2009.
 
Angus DC, Kelley MA, Schmitz RJ, White A, Popovich J Jr. Committee on Manpower for Pulmonary and Critical Care Societies (COMPACCS) Committee on Manpower for Pulmonary and Critical Care Societies (COMPACCS) Caring for the critically ill patient. Current and projected workforce requirements for care of the critically ill and patients with pulmonary disease: can we meet the requirements of an aging population? JAMA. 2000;28421:2762-2770. [CrossRef] [PubMed]
 
The ACGME’s approach to limit resident duty hours 2007-08: a summary of achievements for the fifth year under the common requirements, 2007-2008. Accreditation Council for Graduate Medical Education Web site.http://www.acgme.org/acWebsite/dutyHours/dh_achievesum0708.pdf. Accessed November 18, 2009.
 
Iglehart JK. Revisiting duty-hour limits—IOM recommendations for patient safety and resident education. N Engl J Med. 2008;35925:2633-2635. [CrossRef] [PubMed]
 
Heisler M. Hospitalists and intensivists: partners in caring for the critically ill—the time has come. J Hosp Med. 2010;51:1-3. [CrossRef] [PubMed]
 
D’Agostino R, Halpern NA. Acute care nurse practitioners in oncologic critical care: the memorial Sloan-Kettering cancer center experience. Crit Care Clin. 2010;261:207-217. [CrossRef] [PubMed]
 
Resnick AS, Todd BA, Mullen JL, Morris JB. How do surgical residents and non-physician practitioners play together in the sandbox? Curr Surg. 2006;632:155-164. [CrossRef] [PubMed]
 
Stahlfeld KR, Robinson JM, Burton EC. What do physician extenders in a general surgery residency really do? J Surg Educ. 2008;655:354-358. [CrossRef] [PubMed]
 
Christmas AB, Reynolds J, Hodges S, et al. Physician extenders impact trauma systems. J Trauma. 2005;585:917-920. [CrossRef] [PubMed]
 
Oswanski MF, Sharma OP, Raj SS. Comparative review of use of physician assistants in a level I trauma center. Am Surg. 2004;703:272-279. [PubMed]
 
Victorino GP, Organ CH Jr. Physician assistant influence on surgery residents. Arch Surg. 2003;1389:971-975. [CrossRef] [PubMed]
 
Lundberg S, Wali S, Thomas P, Cope D. Attaining resident duty hours compliance: the acute care nurse practitioners program at Olive View-UCLA Medical Center. Acad Med. 2006;8112:1021-1025. [CrossRef] [PubMed]
 
DeNicola L, Kleid D, Brink L, et al. Use of pediatric physician extenders in pediatric and neonatal intensive care units. Crit Care Med. 1994;2211:1856-1864. [PubMed]
 
Cramer CL, Orlowski JP, DeNicola LK. Pediatric intensivist extenders in the pediatric ICU. Pediatr Clin North Am. 2008;553:687-708. [CrossRef] [PubMed]
 
Kleinpell RM, Ely EW, Grabenkort R. Nurse practitioners and physician assistants in the intensive care unit: an evidence-based review. Crit Care Med. 2008;3610:2888-2897. [CrossRef] [PubMed]
 
Dubaybo BA, Samson MK, Carlson RW. The role of physician-assistants in critical care units. Chest. 1991;991:89-91. [CrossRef] [PubMed]
 
Higgins TL, Teres D, Copes WS, Nathanson BH, Stark M, Kramer AA. Assessing contemporary intensive care unit outcome: an updated Mortality Probability Admission Model (MPM0-III). Crit Care Med. 2007;353:827-835. [CrossRef] [PubMed]
 
Pronovost PJ, Angus DC, Dorman T, Robinson KA, Dremsizov TT, Young TL. Physician staffing patterns and clinical outcomes in critically ill patients: a systematic review. JAMA. 2002;28817:2151-2162. [CrossRef] [PubMed]
 
Halpern NA, Pastores SM. Critical care medicine in the United States 2000-2005: an analysis of bed numbers, occupancy rates, payer mix, and costs. Crit Care Med. 2010;381:65-71. [CrossRef] [PubMed]
 
Angus DC, Shorr AF, White A, Dremsizov TT, Schmitz RJ, Kelley MA. Committee on Manpower for Pulmonary and Critical Care Societies (COMPACCS) Committee on Manpower for Pulmonary and Critical Care Societies (COMPACCS) Critical care delivery in the United States: distribution of services and compliance with Leapfrog recommendations. Crit Care Med. 2006;344:1016-1024. [CrossRef] [PubMed]
 
Carzoli RP, Martinez-Cruz M, Cuevas LL, Murphy S, Chiu T. Comparison of neonatal nurse practitioners, physician assistants, and residents in the neonatal intensive care unit. Arch Pediatr Adolesc Med. 1994;14812:1271-1276. [CrossRef] [PubMed]
 
Mitchell-DiCenso A, Guyatt G, Marrin M, et al. A controlled trial of nurse practitioners in neonatal intensive care. Pediatrics. 1996;986 pt 1:1143-1148. [PubMed]
 
Hoffman L, Tasota F, Zullo T, Scharfenberg C, Donahoe MP. Outcomes of care managed by an acute care nurse practitioner/attending physician team in a subacute medical intensive care unit. Am J Crit Care. 2005;142:121-130. [PubMed]
 
Hoffman LA, Miller TH, Zullo TG, Donahoe MP. Comparison of 2 models for managing tracheotomized patients in a subacute medical intensive care unit. Respir Care. 2006;5111:1230-1236. [PubMed]
 
Hoffman LA, Tasota FJ, Scharfenberg C, Zullo TG, Donahoe MP. Management of patients in the intensive care unit: comparison via work sampling analysis of an acute care nurse practitioner and physicians in training. Am J Crit Care. 2003;125:436-443. [PubMed]
 
Zimmerman JE, Kramer AA, McNair DS, Malila FM. Acute Physiology and Chronic Health Evaluation (APACHE) IV: hospital mortality assessment for today’s critically ill patients. Crit Care Med. 2006;345:1297-1310. [CrossRef] [PubMed]
 
Moran JL, Bristow P, Solomon PJ, George C, Hart GK. Australian and New Zealand Intensive Care Society Database Management Committee (ADMC) Australian and New Zealand Intensive Care Society Database Management Committee (ADMC) Mortality and length-of-stay outcomes, 1993-2003, in the binational Australian and New Zealand intensive care adult patient database. Crit Care Med. 2008;361:46-61. [CrossRef] [PubMed]
 
Kahn JM, Hall JB. More doctors to the rescue in the intensive care unit: a cautionary note. Am J Respir Crit Care Med. 2010;18111:1160-1161. [CrossRef] [PubMed]
 
Gajic O, Afessa B, Hanson AC, et al. Effect of 24-hour mandatory versus on-demand critical care specialist presence on quality of care and family and provider satisfaction in the intensive care unit of a teaching hospital. Crit Care Med. 2008;361:36-44. [CrossRef] [PubMed]
 
Blunt MC, Burchett KR. Out-of-hours consultant cover and case-mix-adjusted mortality in intensive care. Lancet. 2000;3569231:735-736. [CrossRef] [PubMed]
 
Arabi Y, Alshimemeri A, Taher S. Weekend and weeknight admissions have the same outcome of weekday admissions to an intensive care unit with onsite intensivist coverage. Crit Care Med. 2006;343:605-611. [CrossRef] [PubMed]
 
Wunsch H, Mapstone J, Brady T, Hanks R, Rowan K. Hospital mortality associated with day and time of admission to intensive care units. Intensive Care Med. 2004;305:895-901. [CrossRef] [PubMed]
 
Luyt CE, Combes A, Aegerter P, et al. Mortality among patients admitted to intensive care units during weekday day shifts compared with “off” hours. Crit Care Med. 2007;351:3-11. [CrossRef] [PubMed]
 
Laupland KB, Shahpori R, Kirkpatrick AW, Stelfox HT. Hospital mortality among adults admitted to and discharged from intensive care on weekends and evenings. J Crit Care. 2008;233:317-324. [CrossRef] [PubMed]
 
Morales IJ, Peters SG, Afessa B. Hospital mortality rate and length of stay in patients admitted at night to the intensive care unit. Crit Care Med. 2003;313:858-863. [CrossRef] [PubMed]
 
NOTE:
Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging & repositioning the boxes below.

Find Similar Articles
CHEST Journal Articles
PubMed Articles
  • CHEST Journal
    Print ISSN: 0012-3692
    Online ISSN: 1931-3543