Advanced Practice Nurse Outcomes 1990–2008

A Systematic Review

Robin P. Newhouse, PhD, RN, NEA-BC; Julie Stanik-Hutt, PhD, ACNP, CCNS, FAAN; Kathleen M. White, PhD, RN, NEA-BC, FAAN; Meg Johantgen, PhD, RN; Eric B. Bass, MD, MPH; George Zangaro, PhD, RN; Renee F. Wilson, MS; Lily Fountain, MS, CNM, RN; Donald M. Steinwachs, PhD; Lou Heindel, DNP, CRNA; Jonathan P. Weiner, PhD

Disclosures

Nurs Econ. 2011;29(5):230-250.

Abstract and Introduction

Introduction

Quality, access, and cost of health care are high-priority global concerns. In the United States, these issues are pressing due to the escalating cost of managing chronic diseases (Department of Health and Human Services, 2009), the variation in quality of care delivered (Kuehn, 2009), and the inadequate number of primary care physicians (Freed & Stockman, 2009; Kuehn, 2009; Lakhan & Laird, 2009). At this critical time, we still do not know which models of care are best, how to integrate advanced practice registered nurses (APRN) providers, or to what extent APRN providers can contribute to improved access to and quality of health care. These deficits are untenable when the health care needs of society are great and the health reform debate progresses in legislative arenas. How to expand health care services for the American public, at an affordable cost, is central to this dispute.

Advanced practice registered nurses have assumed an increasing role as providers in the health care system, particularly for underserved populations. APRNs complete specialty-specific graduate programs that include education, training, and practice experience needed to complete a national board certification examination before entry into practice. Nurses practicing in APRN roles include nurse practitioners (NPs), clinical nurse specialists (CNSs), certified nurse-midwives (CNMs), and certified registered nurse anesthetists (CRNAs). Several systematic reviews have assessed what is known about NP practice (Brown & Grimes, 1995; Horrocks, Anderson, & Salisbury, 2002; Laurant et al., 2005; Sox, 1979). Similar or better outcomes are found for patient satisfaction (Brown & Grimes, 1995; Horrocks et al., 2002; Laurant et al., 2005; Sox, 1979), patient health status (Horrocks et al., 2002; Laurant et al., 2005), functional status (Brown & Grimes, 1995), and the use of the emergency department (Brown & Grimes, 1995; Laurant et al., 2005). A Cochrane review indicated midwifery care outside the United States was associated with a reduced risk of losing a baby before 24 weeks, a reduced use of regional analgesia, fewer episiotomies or instrumental births, increased chance of a spontaneous vaginal birth, and increased initiation of breastfeeding (Hatem, Sandall, Devane, Soltani, & Gates, 2008). No systematic reviews of CNS or CRNA outcomes have been published.

Although these reviews provide some information about the effects of APRNs on specific outcomes, an updated comprehensive review of the scientific literature on the care provided by APRNs in the United States is needed to inform educational, public, and organizational policy. This review is the most current and complete assessment of the comparability of APRNs to other providers, strengthening and extending the conclusions drawn from previous reviews by including evidence from over a span of 18 years on all types of APRNs and all outcomes, patient populations, and settings.

This systematic review compared the processes and outcomes of care delivered by APRNs to a comparison provider group, most often physicians. The intent was to consider the broad range of studies and outcome measures across these groups using a systematic, transparent, and reproducible review process.

Aim. The aim of this systematic review was to answer the following question: Compared to other providers (physicians or teams without APRNs), are APRN patient outcomes of care similar?

1 2 3 4 5 6 7 8 9

Next Section

Table 1. Quality Assessment Criteria
Criteria	Rating Scale
Were participants in both groups similar?	No (0) Yes (1)
Was setting of both groups similar?	No (0) Yes (1)
Was sample size in both groups adequate?	Less than 30 per group (0) 31–60 per group (1) >60 per group (2)
Were measures reliable and valid?	No (0) Yes (1)
Was bias controlled?	No (0) Yes (1)
Can the outcome be attributed to the APRN?	Yes (2) Partial (1) No (0)
Potential range	0–8

Table 2. Assessment of Overall Evidence
Criteria	Rating System	Definition
Based on the number of studies and numbers of patients, is this sparse?	-1	Sparse = fewer than three studies per outcome; fewer than two RCTs when RCTs are appropriate
As a body of evidence, are the study designs the strongest designs to answer the question?	-1	Determination of strongest study designs is outcome dependent. RCTs are not always feasible, and in some instances, observational studies provide better evidence (e.g., RCT for physiologic outcome such as blood pressure, lipids, glucose — RCT desirable; outcomes that are rare events, such as death, complications — observational desirable).
Is the quality of the studies acceptable?	-1	Quality refers to the study methods and execution. Quality of studies is reflected in the individual study-quality rating (0–8) and designated as low or high (≥5 = high, ≤4 = low).
Is there important inconsistency across the studies?	-1	Consistency is similar estimates of the effect. Inconsistency is demonstrated through differences in direction of effects and significances of differences across all studies. For outcomes for which equivalent non-significant outcomes are favorable (NP, CNM, CRNA), inconsistencies are present when the significant difference favors the comparison group.
Is there concern about the directness of the evidence?	-1	Directness is the extent to which study participants, measures, and outcomes are similar to the population of interest.
Is there a high probability of reporting bias? This includes publication bias and selective reporting of outcomes.	-1	Probability of reporting bias that would result in more significant differences in comparison groups than actually exist

Table 3. Overall Quality Categories and Definitions
Overall Quality	Definition
High	Further research is very unlikely to change our confidence in the estimate of effect.
Moderate	Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low	Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low	Any estimate of effect is very uncertain.

SOURCE: Atkins et al. (2004).

Table 4a. Summary of Study Characteristics for Nurse Practitioners
Author, Year	Compared Groups	Disease/Condition	Patient Population	Setting	Study Quality
RCTs (n=14)
Becker et al., 2005	NP to MD	Coronary artery disease	African Americans, 30–59 y/o, sibling of probands <60 y/o	Community	High
Büla et al. 1999	GNP + team to MD	Varied	>74 y/o, without cognitive or functional impairment	Community	High
Callahan et al., 2006	GNP + team to MD	Alzheimer's disease	In home with caregiver	Community	High
Counsell et al., 2007	NP + team to MD	Varied	>64 y/o; income <200% of federal poverty level	Community	High
Fanta et al., 2006	PNP + attending MD to resident + attending MD	Condition requiring inpatient trauma care	Children between 2 months and 17 years of age	Inpatient	Low
Krichbaum, 2007	GNP to MD	Hip fracture	>64 y/o with hip fracture repair	Inpatient	Low
Lenz et al., 2004	NP to MD	Varied	Hispanic adults with recent urgent care or ED visit	Community	High
Lenz et al., 2002	NP to MD	Diabetes	Adults; primarily Hispanic; no current health care provider	Community	High
Litaker et al., 2003	NP to MD	Hypertension and diabetes	Adults without complex medical conditions	Community	High
Mundinger et al.,2000	NP to MD	Chronic conditions	Hispanic adults with recent urgent care or ED visit	Community	High
Nelson et al., 1991	PNP to usual ED care	Infectious or emergent condition	Children <8 y/o without chronic illness	Community	High
Paez & Allen, 2006	NP to MD	Coronary artery disease	Adults undergoing revascularization procedure	Community	High
Pioro et al., 2001	NP to MD	Varied medical conditions	18–69 y/o; admitted to general medical units	Inpatient	High
Stuck et al., 1995	GNP to MD	Varied conditions	>74 y/o; living at home without preexisting functional impairment	Community	High
Observational (n=23)
Ahern et al., 2004	NP to MD	Chronic hepatitis C	Adults	Community	Low
Aigner et al., 2004	NP to MD	Chronic diseases	Residents in eight nursing homes	Nursing home	High
Aiken et al., 1993	NP to MD	HIV/AIDS	Adults with HIV/AIDS seen in specialty clinic	Community	Low
Bissinger et al., 1997	NNP to MD	Conditions encountered in low- birthweight infants	Low-birthweight neonates between 500–1,250 grams	Inpatient	High
Borgmeyer et al., 2008	PNP to MD	Asthma	Children admitted to general units with exacerbation of asthma	Inpatient	Low
Dahle et al., 1998	NP to MD	Uncomplicated decompensated heart failure	Adults admitted to hospital	Inpatient	High
Garrard et al., 1990	NP to MD usual care	Varied	Nursing home resident Nursing	Home	Low
Gracias et al., 2008	ACNP to MD	Varied postoperative	Adults admitted to surgical ICU	Inpatient	High
Hoffman et al., 2005	ACNP to MD	Varied medical conditions	Adults receiving mechanical ventilation and admitted to subacute MICU	Inpatient	High
Kane et al., 2004	NP to MD	Varied	Nursing home residents Nursing	Home	Low
Karlowicz & McMurray, 2000	NNP to MD	Extremely low-birthweight infants	Newborns admitted to NICU	Inpatient	High
Kutzleb & Reiner, 2006	NP to MD	Heart failure	18–75 y/o	Community	Low
Lambing et al., 2004	NP to MD	Varied	60+ years admitted to geriatric unit or 1 of 2 general medical units	Inpatient	Low
McMullen et al., 2001	ACNP to MD	Varied	Adults admitted to medical unit	Inpatient	High
Meyer & Miers, 2005	ACNP to MD	Varied conditions associated with cardiovascular surgery	Adults undergoing cardiovascular surgery	Inpatient	High
Miller, 1997	GNP to PA	Varied inpatient medical conditions	Nursing home patients admitted to inpatient medicine unit	Inpatient	High
Paul, 2000	NP to MD	Heart failure	Adults with CHF seen in specialty clinic	Community	Low
Pinkerton & Bush, 2000	NP to MD	Diabetes and hypertension	Adults	Community	High
Rideout, 2007	PNP to MD	Cystic fibrosis	Children admitted to adolescent unit	Inpatient	Low
Ruiz et al., 2001	NP to MD	Twin pregnancy	Women diagnosed with twin pregnancy	Community	High
Russell et al., 2002	ACNP to MD	Varied neurological conditions	Adults with tracheostomy admitted to neurological unit	Inpatient	High
Schultz et al., 1994	NNP to MD	Varied neonatal conditions	Infants admitted to transitional care unit	Inpatient	High
Varughese et al., 2006	NP to MD	Varied operative conditions	Children scheduled for outpatient surgery	Ambulatory surgery	Low

Table 4b. Summary of Study Design, Study Groups, Study Purpose, Patient Population, Outcomes, and Quality for Certified Nurse-Midwives
Author, Year	Compared Groups	Disease/Condition	Patient Population	Setting	Study Quality
RCTs (n=2)
Chambliss et al., 1992	CNM vs. MD	Low-risk pregnant women	Admitted in one hospital to unit for physician or unit for midwifery management	Inpatient	High
Heins et al., 1990	CNM to MD	Pregnant women	Women attending 1 of 5 state-funded prenatal clinics and considered high risk for low birthweight	Prenatal-inpatient	High
Observational (n=19)
Baruffi et al., 1990	CNM to MD	Pregnant women	Delivering in hospital with CNM care or hospital with residents and physicians	Inpatient	High
Blanchette, 1995	CNM to MD	Pregnant women	Women in single clinic cared for by CNM or by MD	Prenatal-inpatient	High
Butler et al.,1993	CNM to MD	Low-risk pregnant women	Women delivered in hospital with CNM or MD care	Inpatient	High
Cragin, 2002	CNM to MD	Moderate-risk pregnant women	Women with prenatal care at two sites	Prenatal-inpatient	High
Cragin & Kennedy, 2006	CNM to MD	Low or moderate-risk pregnant women	Women enrolled in obstetric practice who chose either CNM or MD care	Prenatal-inpatient	High
Davis et al., 1994	CNM to MD	Low-risk pregnant women	Women delivering in hospital with both CNM and MD care	Inpatient	High
DeLano et al., 1997	CNM to MD	Low to moderate-risk pregnant women.	Women delivering in hospital with both CNM and MD	Inpatient	High
Fischler & Harvey, 1995	CNM to MD	Retrospective 20-year trend study.	Women cared for in one of three care models and Medicaid, giving birth in single county	Prenatal-inpatient	Low High
Hueston & Rudy, 1993	CNM to MD	Low-income pregnant women	Women receiving care at medical center primary care group	Prenatal-inpatient	High
Jackson, Lang, Ecker et al.,2003a§	CNM to MD	Pregnant women	Women enrolling at study site	Prenatal-inpatient	High
Jackson, Lang, Swartz et al., 2003b§	CNM to MD	Low-income and lowrisk pregnant women	Women enrolling at study site	Prenatal-inpatient	High
Lenaway et al.,1998	CNM to MD	Low-income and lowrisk pregnant women	Women presenting to county health department where CNMs are principal providers	Prenatal-inpatient	High
Low et al., 2000‖	CNM to MD	Low-income pregnant women	Women voluntarily chose the CNM group practice or faculty obstetricians	Prenatal-inpatient	High
MacDorman & Singh, 1998	CNM to MD	Low-risk pregnant women	Women in linked birth/death data	Inpatient	High
Oakley et al., 1995‖	CNM to MD	Low-risk pregnant women	Women in one medical center	Inpatient	High
Oakley et al., 1996‖	CNM to MD	Low-risk pregnant women	Women in one medical center	Inpatient to home	High
Robinson et al., 2000	CNM to MD	Low-risk pregnant women	Women entering care at study site	Inpatient	High
Rosenblatt et al., 1997	CNM to MD	Low-risk pregnant women	Women initiated care with one of the randomly selected providers	Prenatal to inpatient	High
Sze et al., 2008	CNM to MD	Low-risk pregnant women	Women cared for in single hospital	Inpatient	High

§ Represents same study data
‖ Represents same study data

Table 4c. Summary of Study Design, Study Groups, Study Purpose, Patient Population, Outcomes, and Quality for Clinical Nurse Specialists
Author, Year	Compared Groups	Disease/Condition	Patient Population	Setting	Study Quality
RCTs (n=4)
Allen et al., 2002	CNS to usual care	Stroke and transient ischemic attack	Adults discharged to home from hospital or rehabilitation	Community	High
Duffy-Durnin & Campbell-Heider, 1994	CNS to usual care	Medical-surgical admission	>70 y/o	Hospital	High
Swindle et al., 2003	CNS to usual care	Depression	Adult veterans	Community	High
York et al., 1997	CNS to usual care	High-risk pregnancy	Women recruited during first hospital admission	Community	High
Observational (n=7)
Ahrens et al., 2003	CNS to standard care	End-of-life care	Patients at high risk of dying	Inpatient	High
Hanneman et al., 1993	CNS to usual care	Varied conditions	Adults admitted to critical care units	Inpatient	High
Koch & Smith, 1994	CNS to usual care	Radical prostatectomy	Men undergoing retropubic prostatectomy	Inpatient	Low
Lombness, 1994	CNS to PA	Coronary bypass graft	Adults for elective surgery	Inpatient	High
Micheels et al., 1995	CNS to usual care	Colon resection	Adults	Inpatient	Low
Sherman & Johnson, 1994	CNS to usual care	Oncology conditions	Adult oncology patients	Inpatient	Low
Wheeler, 2000	CNS to usual care	Total knee replacement	Adults	Inpatient	High

Table 5a. Summary of Outcomes and Evidence for Nurse Practitioners
Outcome	Number of Studies	Author, Year (Study Quality Rating), Significance	Synthesis of Studies	Evidence Grade
Patient satisfaction	6 (4 RCTs)	Lenz et al., 2004 (6)* Fanta et al., 2006 (3)* Litaker et al., 2003 (8)* Mundinger et al., 2000 (8)* Pinkerton & Bush, 2000 (7) Varughese et al., 2006 (2)	Six studies reported patient satisfaction with the provider. Four of the studies were of high quality (Lenz et al., 2004; Litaker et al., 2003; Mundinger et al., 2000; Pinkerton & Bush, 2000). Five studies were conducted in primary care settings with adults (Lenz et al., 2004; Litaker et al., 2003; Mundinger et al., 2000; Pinkerton & Bush, 2000). The other two studies collected data from parents of children who had undergone outpatient surgery or been admitted to the hospital after a traumatic injury (Fanta et al., 2006; Varughese et al., 2006). When comparing NP and MD care, there is a high level of evidence to support equivalent levels of patient satisfaction.	High: Satisfaction is equivalent in NP and MD comparison groups.
Self-reported perceived health	7 (5 RCTs)	Counsell et al., 2007 (7)^† Litaker et al., 2003 (8) Lenz et al., 2002 (6)* Pioro et al., 2001 (5)* Mundinger et al., 2000 (8)* Ahern et al., 2004 (3) McMullen et al., 2001 (4)†	All used the SF-12 or SF-36 physical and mental function scales to rate self-reported perception of health. Five were judged high-quality RCTs (Counsell et al., 2007; Litaker et al., 2003; Lenz et al., 2002; Mundinger et al., 2000; Pioro et al., 2001). Four of the studies were conducted with adults cared for in a primary care setting (Lenz et al., 2002; Litaker et al., 2003; Mundinger et al., 2000) and one used a sample of adults diagnosed with hepatitis C managed in a specialty clinic (Ahern et al., 2004). A sixth study collected data from older adults receiving home care in a community setting (Counsell et al., 2007). The last two studies reported on results obtained from adults hospitalized with general medical conditions (McMullen et al., 2001; Pioro et al., 2001). One RCT (Counsell et al., 2007) found higher health status in patients cared for by NPs as part of a comprehensive care management team, and the rest of the studies did not find any difference in health status depending on provider type, though two were powered to do so. When comparing NP and MD care, there is a high level of evidence to support equivalent levels of self-reported patient perception of health status.	High: Self-assessed health status is equivalent in NP and MD comparison groups.
Functional Status ADL/IADL	10 (6 RCTs)	Counsell et al., 2007 (7)* Krichbaum, 2007 (3)* Callahan et al., 2006 (5)* Pioro et al., 2001 (5)* Büla et al., 1999 (5)^† Stuck et al., 1995 (8)^† Kutzleb & Reiner, 2006 (2) Aiken et al., 1993 (2) Ahern et al., 2004 (3) Garrard et al., 1990 (3)	Ten studies evaluated the impact of provider (NP vs. MD) on patient functional status in terms of scores on measures of ADL or IADL, 6-minute walk test, or patient self-report. Five of the studies were high quality (Büla et al., 1999; Callahan et al., 2006; Counsell et al., 2007; Pioro et al., 2001; Stuck et al., 1995) and two found NP care was associated with higher functional status (Büla et al., 1999; Stuck et al., 1995). Community-dwelling elders who were recently discharged from hospitals and receiving either home care or inpatient rehabilitation were the focus of five of these studies (Büla et al., 1999; Callahan et al., 2006; Counsell et al., 2007; Krichbaum, 2007; Stuck et al., 1995). One study included adults hospitalized for general medical problems (Pioro et al., 2001) and another included ambulatory patients diagnosed with HIV/AIDS (Aiken et al., 1993). When comparing NP and MD groups, there is a high level of evidence to support equivalent levels of patient functional status.	High: Functional status measured as ADL/IADL is equivalent in NP and MD comparison groups.
Glucose control	5 (5 RCTs)	Becker et al., 2005 (5)^† Lenz et al., 2004 (6) Litaker et al., 2003 (8)^† Lenz et al., 2002 (6) Mundinger et al., 2000 (8)*	Blood glucose control (glycosolated hemoglobin, serum glucose) was an outcome in four studies, all high-quality RCTs. All of the studies were conducted in ambulatory primary care settings using samples of adults (Lenz et al., 2004; Lenz et al., 2002; Litaker et al., 2003; Mundinger et al., 2000). When comparing NP and MD care, there is a high level of evidence to support equivalent levels of patient glucose control.	High: Blood glucose levels/control among patients cared for by NPs was comparable or better than that of patients cared for by other providers
Lipid control	3 (3 RCTs)	Paez & Allen, 2006 (8)^† Becker et al., 2005 (5)^† Litaker et al., 2003 (8)^*†	Three studies examined the effect of provider on serum lipids. All of the studies were conducted in ambulatory primary care settings using samples of adults (Becker et al., 2005; Litaker et al., 2003; Paez & Allen, 2006). The three RCTs were high quality and also provided evidence NP care was associated with better lipid control compared to care from other providers (Paez & Allen, 2006). When comparing NP and MD groups, there is a high level of evidence to support better management of patient serum lipid levels by NPs (Becker et al., 2005; Litaker et al., 2003).	High: Serum lipid levels/control among patients cared for by NP group was better than the MD compari son group.
Blood Pressure	4 (4 RCTs)	Becker et al., 2005 (5)^† Lenz et al., 2004 (5) Litaker et al., 2003 (8)* Mundinger et al., 2000 (8)^*†	Blood pressure control was an outcome of four RCTs. All of the studies were conducted in ambulatory primary care settings using samples of adults. All four RCTs were high quality, and two of those RCTs found patients cared for by the NP had better-controlled BP than patients cared for by other providers (Becker et al., 2005). When comparing NP and MD groups, there is a high level of evidence to support equivalent levels of BP control.	High: Blood pressure levels/control among patients is equivalent in NP and MD comparison groups
ED or urgent care visits	5 (3 RCTs)	Counsell et al., 2007 (7)^† Lenz et al., 2002 (6) Nelson et al., 1991(7)* Aigner et al., 2004 (4) Paul, 2000 (3)	Five studies reported rates of ED visits. All three RCTs were judged to be high quality (Counsell et al., 2007; Lenz et al., 2002; Nelson et al. 1991). Study samples included ambulatory patients with diabetes, hypertension, dyslipidemia, asthma, and heart failure (Lenz et al., 2002; Paul, 2000); community-dwelling elders and nursing home residents (Aigner et al., 2004; Counsell et al., 2007); and otherwise healthy children who had recently been seen in the ED for an emergent condition (Nelson et al., 1991). When comparing NP and MD groups, there is a high level of evidence to support equivalent rates of ED visits.	High: Rates of ED or urgent care visits are equivalent in NP and MD comparison groups.
Hospitalization	11 (3 RCTs)	Counsell et al., 2007 (7)* Stuck et al., 1995 (8)* Lenz et al., 2002 (6)* Schultz et al., 1994 (6) Lambing et al., 2004 (4) Kane, 2004 (4)^† Aigner et al., 2004 (5) Paul, 2000 (4)^† Dahle et al., 1998 (5) Garrard et al., 1990 (3)^† Borgmeyer et al., 2008 (4)	Eleven studies reported rates of hospitalization. Adult patients with heart failure, managed in ambulatory care settings, were the focus of one study (Paul, 2000). Three studies evaluated older adults receiving care in nursing homes (Aigner et al., 2004; Garrard et al., 1990; Kane et al., 2004). The remaining five studies collected data from a variety of individuals discharged home after acute care hospitalizations (premature infants, children with asthma, adults with heart failure, and older adults with general medical conditions) (Borgmeyer et al., 2008; Dahle et al., 1998; Lambing et al., 2004; Schultz et al., 1994). When comparing NP and MD groups, there is a high level of evidence to support equivalent rates of hospitalization.	High: Rates of hospitalization/rehospitalization are equivalent in NP and MD comparison groups.
Duration of ventilation	3 (0 RCTs)	Hoffman et al., 2005 (7) Russell et al., 2002 (5) Bissinger et al., 199(5)	Duration of ventilation was an outcome in three studies. Two found the substitution of an NP for pulmonary fellows and neurosurgical house staff had no deleterious effect on patient duration of ventilation (Hoffman et al., 2005; Russell et al., 2002). Low-birthweight neonates whose care was provided by a neonatal NP or medical residents spent similar lengths of time supported by mechanical ventilation (Bissinger et al., 1997). When comparing NP and MD groups, there is a low level of evidence to support equivalent duration of mechanical ventilation.	Low: Duration of ventilation is comparable among patients cared for by NPs in collaboration with attending MDs compared to duration of ventilation in patients cared for by house - staff MDs in collaboration with attending MDs
LOS	16 (2 RCTs)	Fanta et al., 2006 (3)^† Pioro et al., 2001 (5) Rideout, 2007 (3) Meyer & Miers, 2005 (6)^† Hoffman et al., 2005 (7) Ruiz, 2001 (5)^† Karlowicz & McMurray, 2000 (5) Miller, 1997 (5)^† Schultz et al., 1994 (6)^† Borgmeyer et al., 2008 (4) Lambing et al., 2004 (4)^‡ Aigner et al., 2004 (5) Russell et al., 2002 (5)^† Paul, 2000 (4) Dahle et al., 1998 (5) Bissinger et al., 1997 (5)	High-risk neonates, children (admitted for exacerbation of asthma, pulmonary complications of cystic fibrosis, or non-thoracic or CNS traumatic injuries), adults (admitted with general medical problems or for cardiovascular surgery), and older adults (admitted from home or a nursing home with general medical problems) were included in these studies. In addition, two studies examined outcomes in critically ill adults requiring endotracheal intubation or tracheostomy and mechanical ventilation for respiratory failure. One study was conducted in a neonatal critical care unit with high-risk newborns (excluding those with congenital malformations). Ten were judged high quality. Results of five of the studies favored the NP (Fanta et al., 2006; Miller, 1997; Ruiz et al., 2001; Russell et al., 2002; Schultz et al., 1994) but one low-quality study favored MDs (Lambing et al., 2004). However, the elderly patients cared for by the NPs in that study had higher acuity scores than patients in the MD group. This difference in acuity may have influenced the subsequent patient LOS. Studies in which NP patients had lower LOS included neurosurgical patients, elders, pediatric trauma patients, and low-birthweight and twin neonates. Ten studies found no difference in LOS depending on the provider (NP outcome comparable to physicians). These studies included adults and elderly patients hospitalized in a subacute MICU, cardiovascular surgical patients, and adults diagnosed with a variety of diagnoses, including heart failure, in addition to low-birthweight neonates and children with acute exacerbations of asthma and cystic fibrosis. When comparing NP and MD groups, there is a moderate level of evidence to support equivalent LOS.	Moderate: LOS is equivalent in NP and MD comparison groups
Mortality	8 (1 RCT)	Pioro et al., 2001 (5)* Hoffman et al., 2005 (7) Ruiz, 2001 (5) Karlowicz & McMurray, 2000 (5) Gracias et al., 2008 (7)^† Kane, 2004 (4) Russell et al., 2002 (5) Bissinger et al., 1997 (5)	Samples included high-risk infants (twins, pre-term, or low birthweight) (Bissinger et al., 1997; Karlowicz & McMurray, 2000; Ruiz et al., 2001), adults with acute and chronic medical conditions (Pioro et al., 2001), older adult residents of nursing homes (Kane et al., 2004), and critically ill adults (diagnosed with respiratory failure, multiple-cause critical illnesses, and after complex neurosurgery) (Gracias et al., 2008; Hoffman et al., 2005; Russell et al., 2002). Seven of the studies were judged high quality (Bissinger et al., 1997; Gracias et al., 2008; Hoffman et al., 2005; Karlowicz & McMurray, 2000; Pioro et al., 2001; Ruiz et al., 2001; Russell et al., 2002). A high-quality quasi-experimental study found mortality rates were lower in patients cared for by NPs (Gracias et al., 2008). The remaining seven studies found no differences in mortality rates. When comparing NP and MD groups, there is a high level of evidence to support equivalent mortality rates.	High: Mortality is equivalent in NP and MD comparison groups.

* RCT
^† Favors APRN
^‡ Favors comparison group

Table 5b. Summary of Outcomes for Certified Nurse-Midwives
Outcome	Number of Studies	Author, Year (Study Quality Rating), Significance	Synthesis of Studies	Evidence Grade
Cesarean	15 (1 RCT)	Baruffi et al., 1990 (6)^† Blanchette, 1995 (5)^† Butler et al., 1993 (6)^† Chambliss et al., 1992 (7)* Cragin, 2002 (6)^† Cragin et al., 2006 (5)^† Davis et al., 1994 (6)^† DeLano et al., 1997 (5)^† Fischler & Harvey, 1995 (4)^† Hueston & Rudy, 1993 (7)^† Jackson, Lang, Ecker et al., 2003 (5) Jackson, Lang, Swartz et al., 2003 (5)^† Lenaway et al., 1998 (5) Oakley et al., 1995 (6)^† Rosenblatt et al., 1997 (7)	The only RCT did not show a significant difference. The purpose was to determine if the differences in cesarean rates between the CNMs and obstetricians were due to selection bias. However, it should be noted the baseline cesarean section rates were very low: 2% for CNMs and 9% for obstetricians. Thirteen of the 14 observational studies were high quality. Thirteen of the 15 studies favor CNMs, and the others are equivalent. There is a high level of evidence that CNM patients have lower rates of cesarean sections compared to MD patients.	High: Lower rates of cesarean sections for CNMs than other providers.
Low Apgar score	11 (1 RCT)	Blanchette, 1995 (5) Butler et al., 1993 (6) Chambliss et al., 1992 (7)* Davis et al., 1994 (6) Fischler & Harvey, 1995 (4) Hueston & Rudy, 1993 (7) Jackson, Lang, Ecker et al., 2003 (7) Jackson, Lang, Swartz et al., 2003 (5) Lenaway et al., 1998 (5)^† Oakley et al., 1996 (6) Rosenblatt et al., 1997 (7)	The majority of studies measured as Apgar <7. One of these was a RCT (Chambliss et al., 1992) with a quality rating of 7. For the observational studies, nine were high quality and one was low quality. Since equivalent Apgar scores are desirable, having 10 of the 11 studies with non-significant differences and the remaining study favoring the CNM group was considered acceptable. However, it should be noted several studies included deliveries that might be at risk for low Apgar, while others do not, and there was inconsistent use of statistical control. A high level of evidence indicates CNM and MD Apgar scores are comparable.	High: Comparable rates of low Apgar scores between CNM and other provider groups in all studies but one
Epidural	10 (0 RCTs)	Blanchette, 1995 (5)^† Butler et al., 1993 (6) Cragin, 2006 (5)^† Davis et al., 1994 (6)^† Hueston & Rudy, 1993 (7) Jackson, Lang, Swartz et al., 2003 (7)^† Oakley et al., 1995 (6)^† Robinson et al., 2000 (6)^† Rosenblatt, 1997 (7)^† Sze et al., 2008 (6)^†	Nine of the 10 observational studies showed CNMs used less epidural anesthesia. For births in hospitals, women do have access to regional anesthesia (epidural) during labor even when attended by a CNM. Regional anesthesia may not be available in birthing centers. While there was consistency of findings, there were no RCTs, so the evidence of lower or comparable epidural use was graded as moderate.	Moderate: Less epidural use by CNMs than other providers.
Labor augmentation	9 (1 RCT)	Blanchette,1995 (5)^† Chambliss et al.,1992 (7)^*† Davis et al., 1994 (6)^† Hueston & Rudy, 1993 (7)^† Jackson, Lang, Swartz et al., 2003 (7)^† Lenaway et al., 1998 (5)^‡ Oakley et al., 1995 (6) Robinson et al., 2000(6)^† Rosenblatt et al., 1997(7)^†	One observational study that did not favor the CNM (Oakley et al., 1995) was from a single institution. One study comparing a county-level CNM intervention to two control counties favored the women in the control counties (Lenaway et al., 1998). The authors noted this was contrary to published reports and suggested it may be related to differences in risk or differences in obstetrical practices in institutions. Considering the inclusion of an RCT and the consistency of evidence, the evidence of lower rates of labor augmentation for CNM was graded as high.	High: Lower or comparable use of labor augmentation between CNM and other providers.
Labor induction	9 (0 RCTs)	Blanchette, 1995 (5)^† Davis et al., 1994 (6)^† Hueston & Rudy, 1993 (7)^† Jackson, Lang, Swartz et al., 2003 (7)^† Lenaway et al., 1998 (5)^‡ Oakley et al., 1995 (6) Robinson et al., 2000 (6)^† Rosenblatt et al., 1997 (7)^† Sze et al., 2008 (6)	One of the nine studies showed no significant difference. Seven favored the CNM, while one favored the women in the control counties (Lenaway et al., 1998) similar to labor augmentation. Based on the lack of an RCT study and the inconsistency of the findings, the evidence of lower rates of labor induction for CNM was graded as moderate.	Moderate: Comparable or lower rates of labor induction compared to other providers.
Episiotomy	8 (1 RCT)	Blanchette, 1995 (5)^† Chambliss et al., 1992 (7)^*† Hueston & Rudy, 1993 (7)^† Jackson, Lang, Swartz et al.,2003 (7)^† Low et al., 2000 (6)^† Oakley et al., 1995 (6)^† Robinson et al., 2000 (6)^† Rosenblatt et al., 1997 (7)	All studies were rated as high quality and all favored CNMs. Overall, it was concluded there is a high level of evidence to support that episiotomy rates are lower for CNMs than MDs.	High: Episiotomy rates are lower for CNMs than other providers in all studies.
Low birthweight (<2,500 g)	8 (1 RCT)	Blanchette, 1995 (5) Fischler & Fischler, 1995 (4) Heins et al., 1990 (6)* Jackson, Lang, Swartz et al., 2003 (7) Lenaway et al., 1998 (5) MacDorman & Singh, 1998 (6)^† Oakley et al., 1996 (6) Sze et al., 2008 (6)^†	The common measure was proportion with low birthweight (<2,500 GMs). All remaining observational studies were rated as high quality. While six of the studies reported no significant differences in low-birthweight rates, the other two favored CNMs. There is a high level of evidence that there are comparable rates of low birthweight between CNMs and other providers.	High: Comparable rates of low birthweight between CNMs and other providers.
Vaginal operative delivery (forceps, vacuum, or both)	8 (1 RCT)	Blanchette, 1995 (5) Butler et al., 1993 (6)^† Chambliss et al., 1992 (7)* Cragin, 2002 (6)^† Davis et al., 1994 (6)^† DeLano et al., 1997 (5) Oakley et al., 1995 (6)^† Rosenblatt et al., 1997 (7)^†	Eight high-quality studies reported vaginal operative delivery use, including forceps use, vacuum use, or both. The RCT (Chambliss et al., 1992) showed no significant difference in forceps use but was significant for vacuum use. It should be noted the RCT excluded cases with significant maternal or fetal complications. Five of the remaining seven observational studies favored the CNM. The evidence of lower or comparable vaginal operative deliveries among CNMs was graded as high.	High: Lower or comparable vaginal operative deliveries between CNMs and other providers.
Labor analgesia	6 (1 RCT)	Blanchette, 1995 (5) Chambliss et al., 1992 (7)^*† Davis et al., 1994 (6)^† Hueston & Rudy, 1993 (7) Jackson et al., 2003 (7)^† Oakley et al., 1995 (6)^†	Analgesia (narcotic) use during labor was reported in six studies, one of which was an RCT (Chambliss et al., 1992). The RCT and five of the six observational studies favored the CNM. The studies were all rated as high quality. All women have access to analgesia during labor, but some women prefer to use non-pharmacologic approaches to manage pain. There is a high level of evidence there is less analgesia use by CNMs than MDs.	High: Less analgesia use by CNMs than other providers
Perineal lacerations	5 (1 RCT)	Chambliss et al., 1992 (7)^*† Hueston & Rudy, 1993 (7)^† Low et al., 2000 (6)^† Oakley et al., 1996 (6)^† Robinson et al., 2000 (6)^†	All studies favored the CNM. Perineal lacerations are associated with episiotomy use. A Cochrane review comparing routine versus restricted use of episiotomy found restricted use was associated with less-severe perineal trauma, less suturing, and fewer healing complications (Carroli & Belizan, 1999). Overall, it was concluded there is a high level of evidence rates of third and fourth-degree perineal lacerations are lower for CNMs than MDs.	High: Rates of third and fourthdegree perineal lacerations are lower for CNMs than other providers.
Vaginal birth after cesarean (VBAC)	5 (0 RCTs)	Blanchette, 1995 (5)^† Cragin, 2002 (6)^† Davis et al., 1994 (6)^† DeLano et al., 1997 (5)^† Lenaway et al., 1998 (5)	Four of the five studies favored CNMs. The one study that showed no difference (Lenaway et al., 1998) did have a higher proportion but it was not significant in random-effects testing. Not all of the studies excluded women who may not be eligible for VBAC, and there were no RCTs. A moderate level of evidence supports comparable or higher rates of VBAC for CNMs compared to MDs.	Moderate: Comparable or higher rates of VBAC for CNMs compared to other providers.
NICU admission	5 (0 RCTs)	Butler et al., 1993 (6)^† Fischler & Harvey, 1995 (4) Hueston & Rudy, 1993 (7) Jackson, Lang, Swartz et al., 2003 (7) Oakley et al., 1996 (6)^†	Two of the studies (Butler et al., 1993; Oakley et al., 1996) reported lower admission for the CNM group. One of these (Butler et al., 1993) used statistical control for age, race, year of delivery, infant size, and parity. Based on the lack of RCTs and inconsistent results, the evidence supporting lower NICU admission was considered moderate.	Moderate: Comparable or lower rates of NICU admission for CNM compared to other providers
Breastfeeding	3 (0 RCTs)	Cragin, 2002 (6)^† Jackson, Lang, Swartz et al., 2003 (7)^† Oakley et al., 1996 (6)^†	Three observational studies reported differences in initiation of breastfeeding. All were rated as high quality and avored CNMs. There is a moderate level of evidence CNM patients have higher breastfeeding rates than MDs.	Moderate: CNM patients have higher breast feeding rates than other providers.

* RCT
^† Favors APRN
^‡ Favors comparison group

Table 5c. Summary of Outcomes for Clinical Nurse Specialists
Outcome	Number of Studies	Author, Year (Study Quality Rating), Significance	Synthesis of Studies	Evidence Grade
Satisfaction	3 (2 RCTs)	Johnson & Sherman, 1994 (4)^† Swindle et al., 2003 (7)* York et al., 1997 (7)*	Only one low-quality observational study found a significant difference favoring the CNS group in satisfaction with care in a pre-test post-test inpatient oncology CNS case-management study (Johnson & Sherman, 1994). A high level of evidence supports comparable satisfaction scores, indicating that the CNS may not have a direct effect on patient satisfaction.	High: Satisfaction among patients on units with a CNS was similar to comparison group.
Length of stay	7 (2 RCTs)	Ahrens et al., 2003 (6)^† Duffy-Durnin & Campell-Heider, 1994 (5)* Koch & Smith, 1994 (3)^† Lombness, 1994 (6)^† Micheels et al., 1995 (4) Wheeler, 2000 (8)^† York et al., 1997 (7)*	Four of the five observational studies demonstrated lower LOS for patients on units with CNS-enhanced care for populations of post-coronary bypass surgery, patients in end-of-life care, patients undergoing radical prostatectomy, and patients who had a total knee replacement. Two RCTs found no significance difference in LOS when the CNS was involved in the care. However, one of those studies (York et al., 1997) explored CNS in postpartum care, and postpartum LOS is generally fixed, resulting in non-significance. Three high- and one low-quality observational studies demonstrated a difference favoring the CNS group.	High: LOS is comparable or better than patients cared for by a CNS as compared to non-CNS.
Cost	4 (2 RCTs)	Ahrens, 2003 (6)^† Koch & Smith, 1994 (3)^† Swindle et al., 2003 (7)* York et al., 1997 (7)^*†	Three of the studies reported a significantly lower cost; however, the comparisons in each of the three studies were different. The comparisons were between CNS postpartum care and standard care (York et al., 1997) CNS and MD, with MDs in a population of patients at risk to die (Ahrens et al. 2003) and pre-post CNS with guideline implementation (Koch et al., 1994). One RCT reported no difference in the cost of care when utilizing a MD/CNS intervention for patients with major depression in an ambulatory setting (Swindle et al. 2003). A high level of evidence supports utilization of the CNS role decreases costs of care.	High: Cost is lower in CNS group care.
Complications	3 (1 RCT)	Allen et al., 2002 (6)^*† Hanneman et al., 1993 (6) (Medical^†) Hanneman et al., 1993 (6) (Surgical) Lombness, 1994 (6)	In the RCT of post-discharge care management by a team with a CNS, stroke and transient ischemic attack patients experienced lower complications than usual care. (Allen et al., 2002). In one observational study, patients in a surgical intensive care experienced less endotracheal tube malposition and inadvertent extubation (Hanneman et al. 1993). In two observational studies, no difference in complication rates were found for postoperative cardiac surgery patients managed by a PA or CNS (Lombness, 1994), and in pre- and post-surgical patient pulmonary complications (Hanneman et al. 1993). Because of the predominance of the pretest post-test design and inconsistency in results, a moderate level of evidence supports that CNSs affect lower complication rates.	Moderate: Complications are lower or comparable when CNS is involved in care as compared to non-CNS staff.

* RCT
^† Favors APRN
^‡ Favors comparison group

Comments

Commenting is limited to medical professionals. To comment please Log-in.

Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.

processing....

Advanced Practice Nurse Outcomes 1990–2008

A Systematic Review

Abstract and Introduction

Introduction

Tables

Tables

Tables

Tables

Tables

Tables

Tables

Tables

Tables

References

Authors and Disclosures

Authors and Disclosures

Sidebar

Sidebar

Executive Summary

Comments

Email This

Feedback