DoctorFinder | Join/Renew | MyAMA | Site Map | Contact Us

Impact of 24-hour postpartum stay on infant and maternal health

Note: This report represents the medical/scientific literature on this subject as of June 1995.

Full text

Resolution 135 (A-94), introduced by the District of Columbia Delegation, was referred to the Board of Trustees (BOT). Resolution 135 called for the AMA to review existing data or, if such is insufficient, deliver a survey of the impact of insurance-driven discharge of mothers and newborns within 24 hours of delivery with particular emphasis on (1) health of the infant, (2) health of the mother (physical and emotional), (3) issues dealing with instruction/education, i.e., breast feeding, well baby care, immunizations, (4) effects on remainder of family members, and (5) personal and financial effects. In response, Board Report 24-(I-94) directed the Council on Scientific Affairs to study the issue of 24-hour postpartum discharge of mothers and infants, to evaluate the impact of this practice on infant and maternal health, and to report back to the House of Delegates at the 1995 Annual Meeting. 

Background 

A variety of sociological and economic factors have converged to shorten dramatically postpartum stays for most American women and infants. Typical hospital stays used to be four or five days for normal vaginal births and one to two weeks for cesarean births. Over the past decade this decreased, and in the last three or four years typical hospital stays have dwindled to 24 hours or less for uncomplicated vaginal deliveries and two to three days for Cesarean deliveries. The driving force behind these changes has been largely socioeconomic. However, some evidence exists that mothers may prefer shorter stays, and some researchers have concluded that complication rates for abbreviated postpartum stays are no higher than for traditional stays. 

The first 72 hours of life are a time of significant transition and stabilization for newborn infants. It is during this period that a variety of important tasks must take place. The mother goes through a period of rest and recuperation and is further educated about infants, breastfeeding, and parenting. The child is observed for ductal and cardiac abnormalities, respiratory difficulties, infections, jaundice, adequacy of feeding, and gastrointestinal anomalies. Newborn screening occurs, and the mother-child dyad is assessed for evidence of biological and environmental risk factors. If any of these preventive measures do not occur, potential risk to mother or child increases. 

The first issue in examining the question of "early discharge" is attempting to define it. The Guidelines for Perinatal Care, jointly published by the American Academy of Pediatrics (AAP) and the American College of Obstetricians and Gynecologists (ACOG), define early discharge as discharge within 48 hours of birth. It refers to discharge in less than 24 hours as very early discharge.1 

Empirical studies have varied in their definition of early discharge. Most research examines the impact of discharge within 48 hours of vaginal delivery. Most expressions of concern have targeted discharges occurring in substantially less than 24 hours. Consequently, the available literature is limited and often does not focus on the populations of specific concern. Further complicating the analysis is the fact that most studies have had rigorous inclusion and exclusion criteria and have been conducted on mothers requesting early discharge. There is therefore only limited ability to evaluate scientifically the effects of shortening postpartum stays. 

When evaluating the clinical effects of shorter stays, the following considerations are appropriate: 

  • health effects on and complication rates in the infant  
  • health effects (both physical and psychological) on the mother  
  • extent of opportunity for maternal and infant-care education  
  • physical and psychological effects on other family members  
  • extent of opportunity for maternal and child psychosocial assessment. 

Available research 

Studies examining the issue of early discharge have tended to examine model programs that included a pre-screening process to identify low-risk mothers and infants, some degree of prenatal instruction or education, and some degree of follow-up care. The exact nature of these interventions, however, varied from program to program. Follow-up care, for instance, ranged from one home visit and a phone call to three home visits in the first three days home, and telephone consultation available 24 hours a day for two weeks.2 

Methods of participant selection also varied from program to program. In almost all cases, however, the women chose whether they wanted early discharge or traditional hospital stay and the option was only available to a selected population. One study did use a randomized method of participant selection, but the authors acknowledge that "sample sizes were too small to detect significant differences in the outcomes."3 

There are several factors that make interpretation of the available research tentative at best: 

  • discrepant definitions of "early discharge"  
  • exclusion of women and infants likely to have birth-related complications  
  • absence of random assignment  
  • no or flawed control groups  
  • provision of compensatory services in early discharge groups that may mask or  
  • alter the effects of more rapid discharge. 

Type I vs. Type II error 

Another important question is the issue of where the burden of proof should lie. Is it sufficient that a change in practice fail to create significant adverse consequences, or should it be demonstrated to be "safe?" At the heart of this debate is the issue of Type I vs. Type II error--failure to reject the null hypothesis does not mean that the null hypothesis is correct. In other words, evidence that early discharge is not unsafe does not mean that it is safe. 

In a review commissioned by the Maternal and Child Health Bureau (MCHB) of the Health Resources and Services Administration, Paula Braveman concludes that insufficient statistical power is a methodological flaw common to all published studies of early discharge programs--"None of the studies in the literature had adequate power to detect even a doubling of the neonatal readmission rate."4 

Despite this problem, most researchers in the field of early discharge have concluded that early discharge is safe, based on failure to demonstrate significant differences in neonatal readmission rates. 

What is a good outcome? 

One issue that bedevils outcome research is the difficulty in operationalizing a positive outcome. With regard to the issue of early discharge, the outcome variable generally selected is readmission of the infant within the first year of life. Information about morbidities not requiring rehospitalization is generally unavailable. Further complicating the issue is the economically driven trend to treat problems on an outpatient basis as much as possible, which may result in shifting criteria for readmission over the study period. An argument can also be made that readmission in the context of a follow-up visit is a good outcome, in that problems are being diagnosed and treated that might otherwise have been missed. 

Health of the infant 

Within this shifting empirical landscape, a few studies have drawn conclusions about the impact of abbreviated postpartum hospital stays. The preponderance of available literature addresses the health status of infants going home after shorter vs. longer hospital stays. 

Clearly, early discharge minimizes the risk of nosocomial infection (for both infant and mother).5 However, it may not provide adequate time for routine medical and social assessment of the dyad.6 Most researchers have concluded that early discharge (discharge within 48 hours of delivery) is generally safe for low-risk infants, based on a lack of "significant differences in the levels or types of mortality or readmissions for the early discharge and control groups."2 

A review of all postpartum early discharge program outcomes in the United States published between 1960 and 1985 concluded that discharge prior to 48 hours (most within 24 hours) was not associated with an increase in neonatal morbidity.2 None of the programs included in this review, however, had adequate sample sizes for good statistical power, and at least two thirds of the programs lacked a control group.4 A more recent study (1990) found no significant difference in infant morbidity between infants randomly assigned to be discharged after 12-24 hours, 25-48 hours, and four days. However, small sample size limits the utility of these results.3 

Almost all instances of infant readmission (in all studies) were due to hyperbilirubinemia, with variability due to differences in the definition and treatment of hyperbilirubinemia. Complicating this finding is the fact that hospitalization is no longer considered to be the gold standard for treating moderate hyperbilirubinemia in neonates. Norr and Nacion report that "of a total of 84 infant readmissions reported by all of [the] studies, 71 were for hyperbilirubinemia and only 13 were for all other causes, including transient tachypnea, bradycardia, and hypothermia, with no one cause predominating." Norr and Nacion further report that morbidity from causes other than hyperbilirubinemia is low. Superficial skin infections, feeding problems (lack of weight gain), and hypothermia are cited as the most common problems. Less common problems include diarrhea, pneumonia, convulsions, and respiratory difficulties.2 

There are many anecdotal reports about other serious problems such as failure to thrive, missed diagnoses of metabolic disorders such as phenylketonuria, and serious bleeding from circumcision wounds, but these have not as yet been documented empirically. 

Perinatal screening 

Mandated neonatal screening varies from state to state, but all screening programs are designed to prevent morbidity and mortality through early diagnosis of medical problems and congenital disorders that can have serious sequelae. Adequate screening programs require universal participation, prompt diagnosis, and mechanisms for parental notification and education. Common screening tests include those for persistent hyperphenalalaninemias, including phenylketonuria, galactosemia, maple syrup urine disease, congenital adrenal hyperplasia, and congenital hypothyroidism. Some states also mandate screening for sickle cell disease and various sexually transmissible infections. 

Historically, these screening samples have been taken on the second or third day after delivery. This time period was established to optimize the sensitivity of the tests. PKU screening, for example, has a false negative rate that is quite high (about 33 percent) during the first 12 hours of life and drops to less than 1 percent after 48 hours. In contrast, early screening for congenital hypothyroidism, especially using thyrotopin (TSH) as the primary screen, may have a high false positive rate because of the normal TSH surge that accompanies parturition. Some experts argue that screening samples taken in the first 24 hours of life serve merely to notify the system of the existence of the infant. Follow-up screening for PKU or congenital hypothyroidism, for example, must then be completed in a timely fashion, no later than two weeks after birth.1 Loss of infants to follow-up screening is a serious concern, but data regarding the rates of missed rescreening are not yet available. 

Health of the mother 

Methodological flaws also plague the few studies examining the impact of early discharge on the physical health of the mother. Research has shown that very few mothers (less than two percent) require rehospitalization after discharge.2,7 When they do, the problems are often fairly serious, including significant infection and hemorrhaging. The majority of mothers who did require readmission in one review suffered from endometritis.2 In that review, six of the ten readmissions reported in four studies were for endometritis. (Two of the other four women were readmitted for postpartum hemorrhages, one for severe postpartum depression, and one for a severely abscessed episiotomy.)2 Again, however, these findings are not necessarily conclusive. 

The MCHB report points out that these studies had "comparison groups . . . [that] were likely to differ on relevant characteristics . . . [and/or] sample sizes were inadequate."4 The data on maternal morbidity treated on an outpatient basis are even less certain. Many studies do not provide specific information about complication rates; others provide no information at all.2,3 Of the studies that did offer information about morbidity that did not require hospitalization, the most common problems included mild hypertension, anemia, breast or episiotomy problems, and fatigue.2,3,7 In a 1990 study that included a control group, the mothers experienced a generally healthy postpartum course, regardless of the time of discharge.3 

The impact of early discharge on the emotional health of the mother is a similarly important area of concern. Emotional health includes such things as postpartum depression, mother-child separation, the mother's confidence in her ability to be a good parent, stress management, and the need for external support. In the randomized, controlled study that compared discharge at 12-24 hours, 25-48 hours, and four days, the mothers discharged 12-24 hours after giving birth reported themselves to be significantly less depressed at one month postpartum than the two later groups. In addition, the earliest group reported higher levels of confidence at one week postpartum than the other two groups. At one month postpartum, however, the three groups did not differ significantly with respect to levels of confidence. This suggests that women who assume total responsibility for their baby sooner after delivery may feel more confident initially than those who do not, but the direction of causality is unclear.3 

Another study, which compared women who had domiciliary care after early discharge to women who had hospital care until discharge, found that the difference between the emotional health of mothers in both groups was not statistically significant.8 

Although many of the studies mention other aspects of emotional health such as mother-baby bonding and the mother's support system, they do not provide data. A second document commissioned by MCHB reports on an October 1994 consensus meeting in Boston. There, a group of experts with "intimate knowledge of both hospital and community based care for mothers and newborns" found that "women who are from high risk social environments and with limited social support structures may not be best served by a one-day stay." In addition, "separation due to the need to observe either the mother or the newborn was deemed unacceptable . . . . Policies that would result in increased rates of separation were viewed by the group as detrimental to mother infant bonding, infant and maternal health, and threatening to breastfeeding. Clinicians recognized a distinction between 24 hours and 72 hours because mother/infant bonding occurs during this early and critical time period."9 

Education 

Many experts express concern that early discharge will reduce the degree to which health care providers can offer parental education and instruction. The decreased length of hospital stay associated with early discharge significantly increases the importance of comprehensive and continuous prenatal and postpartum care.10 Such care includes breastfeeding instruction, early discharge planning, and educating the mother about immunizations, well-baby care, and bodily changes. 

All of the studies reported in the literature included fairly extensive prenatal and follow-up care. No data are available about the relative educational needs of women who present for delivery without benefit of prenatal care, regardless of length of hospital stay. 

Impact on other family members 

The impact of early discharge on the rest of the family is not as well documented as the impact on the mother and the infant. The studies do cite family togetherness and the involvement of the father in caring for the new baby as sources of parent satisfaction.5,8,11-13 These studies, however, focus on the effect of early discharge on the mother and not on the rest of the family. The existing research does not sufficiently examine the impact of early discharge on other family members. 

Cost-effectiveness 

The cost-effectiveness of early discharge programs is the final consideration examined in the literature. The literature only investigates the cost-effectiveness for the hospital. It does not address the impact on the patient's costs. Some studies conclude that reduced lengths of stay and the resulting increase in the number of beds available for other patients render early discharge programs cost-effective.13,14 Other authors claim that no one has thoroughly analyzed whether early discharge programs truly provide a way to contain costs.12 Again, the existing data are sketchy and inconclusive. 

Conclusions 

We still do not know how much and what kind of follow-up is needed for maternal and infant safety.12 

In the MCHB review, Braveman "categorized studies of early discharge according to the type of compensatory service provided in the early post-discharge days . . . [b]ecause conclusions from studies about early discharge in the presence of a given compensatory service cannot be generalized to early discharge in the absence of that service."4 The reviewed studies that included post-discharge services did indicate that most women were generally satisfied with the care they received for both early discharge and traditional stay. At least two studies (both with postpartum home visiting) report that mothers who were discharged early were more satisfied than mothers who had traditional hospital stay.3,8 Satisfaction, however, depends upon expectations, and studies show that the expectations of women who choose early discharge often differ from those of women who choose a traditional hospital stay. This suggests that women should be allowed to choose the maternity care which best suits their individual needs.8 

While we can therefore conclude that some women prefer to be discharged sooner than later, definitive studies regarding the safety of doing so are not available at this time. The limitations of the existing research are significant. Inadequate statistical power, absence of random assignment, careful participant selection, and unstable definitions of "early" and "good outcome" render scientific conclusions highly unreliable. One of the major limitations of the studies is that most of the early discharge programs were voluntary. Norr and Nacion acknowledge that "it is likely that families who choose early discharge not only have favorable attitudes about coming home early, but also the resources and social support needed to make the discharge go smoothly."1 Second, very few researchers have studied the effects of early discharge on disadvantaged populations -- most participants are either middle-class or upper middle-class, healthy, well-educated, and living in a stable home environment.2,3 The outcomes of the studies, therefore, cannot be generalized to high-risk populations. Further, there is no national consensus on what constitutes "low-risk" versus "high-risk." The October 1994 consensus meeting concluded that "neither the AAP/ACOG Guidelines for Perinatal Care regarding eligibility for 24 hour discharge nor the specifications concerning follow-up seemed adequate from the perspective of clinicians."9 

Some physicians worry that the current trend toward reduced hospital stays will have dire results. Arthur Eidelman, MD, for example, notes that "it is assumed [that] the mother will be discharged into a supportive family unit where another responsible adult is present and where there exists an accessible medical care system for follow-up and ongoing care." Fully aware that many women do not enjoy such luxuries, Eidelman warns that allowing high-risk mothers such as "an adolescent primipara, or a single mother subsisting on welfare . . . into a system that does not have any backup visiting nurse program or community health clinics is simply committing medical and social suicide."4 Eidelman's concerns reinforce the necessity of limiting early discharge to low-risk mothers or providing comprehensive care that would ensure a safe and healthy discharge for high-risk patients. 

The commissioned MCHB review concludes, "Although many studies have examined early discharge, when standard scientific criteria are applied it becomes clear that the currently available literature provides little scientific knowledge to guide discharge planning for apparently well newborns and their mothers. . . . Rigorously designed studies are needed to examine the impact of early discharge and a range of post-discharge practices not only on newborns but on mothers and families. . . .[I]nformation from sound research is needed to inform policy addressing third party reimbursement as well as clinical care standards."4 

In December 1994, the MCHB convened a group that included the American Medical Association and other concerned organizations, academicians, and experts to begin the process of examining the issue of early perinatal discharge. It was agreed that there may be cause for concern in the present trend toward more and more rapid postpartum discharge. Principles agreed upon at that meeting include: 

Discussions of an acceptable vs. unacceptable number of postpartum hours are unproductive and should be replaced by work to identify the essential prerequisites for timely and appropriate perinatal discharge of mother and infant 

Discharge policies that result in separations between mothers and infants due to staggered hospital releases are undesirable 

There needs to be concentrated work on developing national standards for adequate and sufficient care during the early neonatal period 

Recommendations

The following statements, recommended by the Council on Scientific Affairs, were adopted by the AMA House of Delegates as AMA policy at the 1995 AMA Annual meeting: 

  1. The AMA is concerned about the trend toward increasingly brief perinatal hospital stays as a routine practice in the absence of adequate data to demonstrate the practice is safe.  
  2. Policy of the AMA is that in the absence of definitive empirical data, perinatal discharge of mothers and infants should be determined by the clinical judgment of attending physicians and not by economic considerations. This decision should be made based on the criteria of medical stability, delivery of adequate predischarge education, need for neonatal screening, and determination that adequate feeding is occurring, and with consideration of the mother's social and emotional needs and preferences. A plan should be in place for psychosocial and medical follow-up, as outlined in the Guidelines for Perinatal Care developed by the American Academy of Pediatrics and the American College of Obstetricians and Gynecologists.  
  3. The AMA should encourage well-designed experimental studies [that] are needed to identify safe neonatal practices with regard to the hospital discharge of mothers and infants.  
  4. The AMA supports the collaborative efforts of the Maternal and Child Health Bureau and other concerned national organizations to examine more thoroughly the issue of appropriate medical care during the perinatal period.

References 

  1. American Academy of Pediatrics and American College of Obstetrics and Gynecology. Guidelines for Perinatal Care, 3rd Edition. Elk Grove Village, IL, AAP, 1992.  
  2. Norr, K, Nacion, K. Outcomes of postpartum early discharge, 1960-1986: A comparative review. Birth. 1987;14:135-140.  
  3. Carty, EM, Bradley, CF. A randomized, controlled, evaluation of early postpartum hospital discharge. Birth. 1990;17:199-204.  
  4. Braveman, P. Early discharge of newborns and mothers: a critical review of the literature. Unpublished Report commissioned by Maternal and Child Health Bureau, Department of Health and Human Services, December 1994.  
  5. Jansson, P. Early Postpartum Discharge. Am J Nursing. 1985;May:547-550.  
  6. Eidelman, AI. Early discharge - early trouble. J Perinat. 1992;XII:101-102.  
  7. Beck, CT. Early postpartum discharge programs in the United States: A literature review and critique. Women & Health. 1991;17:125-138.  
  8. Kenny, P, King, MT, Cameron, S, Shiell, A. Satisfaction wth postnatal care - the choice of home or hospital. Midwifery. 1993;9:140-153.  
  9. Schwartz, RM. Short stay hospitalization for mother and newborns: concerns and issues. Submitted to Maternal and Child Health Bureau. Providence RI, The National Perinatal Information Center, November 1994.  
  10. Gillerman, H, Beckham, MH. The postpartum early discharge dilemma: An innovative solution. J Perinat Neonatal Nurs. 1991;5:9-17.  
  11. Campbell, IE. Early postpartum discharge - an alternative to traditional hospital care. Midwifery. 1992;8:132-142.  
  12. Lukacs, A. Issues surrounding early postpartum discharge: Effects on the caregiver. J Perinat Neonatal Nurs. 1991;5:33-42.  
  13. Berryman, GK. Early discharge of mothers and infants following vaginal childbirth. Military Medicine. 1991;156:583-584.  
  14. Gonzalves, PE. Coordinated care early discharge of postpartum patients at Irwin Army Community Hospital. Military Medicine. 1993;158:820-822.

Reports by topic

Last updated: Sep 24, 2007
Content provided by: CSAPH