HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Gould, J. B. Articles by Chavez, G. Search for Related Content PubMed PubMed Citation Articles by Gould, J. B. Articles by Chavez, G.

Time of Birth and the Risk of Neonatal Death

From the Division of Neonatal and Developmental Medicine, Stanford University, Palo Alto, California; School of Public Health, University of California, Berkeley, California; and California Department of Health Services, Sacramento, California.

	ABSTRACT

TOP ABSTRACT METHODS RESULTS DISCUSSION REFERENCES

 
Objective: To assess whether mortality is increased in the United States in infants born at night, we compared case-mix adjusted neonatal mortality for low- and high-risk infants born during the daytime (7 am to 6 pm), early night (7 pm to 12 am), and late night (1 am to 6 am).

Methods: California linked birth-death certificate data on 3,363,157 infants, weighing more than 500 g and born without lethal congenital anomalies in 1992–1999, were analyzed. Logistic regression, adjusting for birth weight, gender, prenatal care initiation, maternal hypertension, eclampsia, diabetes, and placental abruption/previa, was used to estimate the relationship between neonatal mortality and time of birth.

Results: The overall neonatal mortality was 2.08 deaths per 1,000 live births. Neonatal mortality was 1.88 for daytime births, increasing to 2.37 for early night and 2.31 for late night births. After adjusting for case mix, early night births had a 12% increase and late night births a 16% increase in the odds of neonatal death, an excess that accounts for 9.6% of all neonatal deaths. Mortality was increased for night births that were less than 1,500 g or more than 1,500 g, singletons or multiples, and those delivered vaginally or by cesarean. The increased risk was identified in hospitals that provide intermediate, community, and regional neonatal intensive care, but not in hospitals that provide primary care.

Conclusion: Identifying the causal factors and reducing the increased burden of mortality for infants born at night should be a major priority for perinatal medicine.

Level of Evidence: III

	METHODS

TOP ABSTRACT METHODS RESULTS DISCUSSION REFERENCES

 
This analysis is based on 3,363,157 live births without lethal congenital anomalies (weight equal to or greater than 500 g) born in California between 1992 and 1997. Data were obtained from the California linked birth-infant death cohort files, which contain information on prenatal care, perinatal complications, birth weight, gender, plurality, method of delivery, time of birth recorded in hours and minutes, and cause of death, coded by International Classification of Diseases, 9th Revision (ICD-9).⁹ An additional 363 records (0.011%) could not be included in the cohort because time of birth was not recorded. These publicly available files do not include personal identifiers. Any infant whose cause of death was recorded as a congenital abnormality (ICD-9 codes 740–7484 and 7486–7599) was excluded from the study cohort. Neonatal mortality was defined as the number of deaths during the first 28 days of life per 1,000 live births.

The nighttime was arbitrarily separated into two 6-hour periods: early night (7 pm to 1 am, 19:00–00:59) and late night (1 am to 7 am, 1:00–6:59). Logistic regressions were used to evaluate the effects of early and late night birth on neonatal death by using daytime (7 am to 6 pm, 7:00–18:59) as the reference. To control for differences in case mix, the trimester of prenatal care initiation, maternal pregnancy/delivery complications (high blood pressure, diabetes, preeclampxia and eclampxia, placenta previa, and abruptio placenta), gender, and birth weight were included in the model. Odds ratios (ORs) and their 95% confidence intervals (CIs) are reported. Separate models were developed for each of the higher-risk (VLBWs, multiples, cesarean) and lower-risk (non-VLBW, singleton, vaginal delivery) subgroups. Because California hospitals are not classified by obstetric level of care, we have used neonatal level of care as designated by California Children's Services. Seventy-four percent of the cohort (2,478,201 infants) were born at hospitals officially designated by California Children's Services as providing primary, intermediate (morbidity not requiring ventilation), community (long-term ventilation, no major surgery), and regional neonatal intensive care (ventilation and major surgery).¹⁰ Separate models were developed to assess the mortality of night births by level of care. Using a modification of the Dollfus classifications,¹¹ the cause of death recorded on the vital record was categorized as asphyxia, obstetric conditions, respiratory distress syndrome (RDS), other respiratory disorders, prematurity, perinatal infection, neonatal infection, and other causes (ICD-9 codes for these categories are available upon request). Logistic models were developed to assess the relationship between time of birth and cause specific death, controlling for prenatal care, perinatal complications, gender, and birth weight. Descriptive analyses and logistic regressions were conducted with SAS 8.1 (SAS Institute Inc, Cary NC). This study was approved by the Stanford University Institutional Review Board.

	RESULTS

TOP ABSTRACT METHODS RESULTS DISCUSSION REFERENCES

 
There were striking temporal trends in the pattern of deliveries. Births were slightly more prevalent during the daytime (56.7% of births), decreasing to 23.4% in the early night and 19.9% during the late night. Although the percentage of multiple births decreased (2.5% in the daytime, 2.3% in early night, and 1.9 % in late night), there was an increase in the percentage of VLBW births (0.89% in the day, 1.07% in early night, and 0.95% in the late night). The percentage of infants delivered by cesarean dramatically decreased from 24.4% during the day to 19.8% during the early evening. During the late evening only 12.7% of deliveries were by cesarean. There were also temporal trends in maternal demographics associated with perinatal risk. The percentages of black non-Hispanics, teenagers, mothers with only third trimester or no prenatal care, and those with hypertension were consistently increased at night (Table 1).

Neonatal mortality was higher for infants born at night (7 pm to 7 am) than for infants born during the day (7 am to 7 pm). Figure 1 shows the relationship between hour of birth and neonatal mortality. The neonatal mortality for infants born during the day was 1.88 per 1,000 live births. Neonatal mortality increased to 2.37 for infants born during the early night and remained elevated at 2.31 for infants born during the late night. After adjusting for adequacy of prenatal care, complications of pregnancy, gender and birth weight, the OR for neonatal mortality was 1.12 (95% CI 1.05–1.19) for infants born in the early night and 1.16 (95% CI 1.09–1.23) for infants born during the late night (Table 2). There is little difference in the extent of disadvantage seen in early night births compared with late night births.

View larger version (27K): [in this window] [in a new window]   Fig. 1. Neonatal mortality by time of birth, California 1992–1997. Dotted line is observed neonatal mortality. Solid line is a 2-point moving average of neonatal mortality. Neonatal mortality = deaths per 1,000 live births in the first 28 days. Data source: California Department of Health Services, Birth Cohort File, 1992–1997. Includes live births without congenital abnormalities occurring in California with birth weight of at least 500 g. Births = 1,905,157 for 7 am to 6 pm (days), 786,715 for 7 pm to 12 am (early night), and 670,911 for 1 am to 6 am (late night). Gould. Time of Birth and Neonatal Mortality. Obstet Gynecol 2005.

Table 2 also shows the neonatal mortality rate by time of birth for infants at higher and lower risk. Both the higher-risk VLBW infants and the lower-risk infants weighing more than 1,500 g demonstrated an increase in the neonatal mortality rate when born during the early night (OR 1.10, 95% CI 1.02–1.19, and OR 1.15, 95% CI 1.04–1.26) and late night (OR 1.18, 95% CI 1.09–1.28, and OR 1.12, 95% CI 1.01–1.24). However, although neonatal mortality rate was significantly elevated for singletons born in the early night (OR 1.11, 95% CI 1.04–1.18) and in the late night (OR 1.16, 95% CI 1.08–1.24), the time effect was less striking for multiples. For multiples, neonatal mortality rate was only elevated for births during the early night (OR in early night 1.17, 95% CI 1.01–1.36, P = .04 versus OR in late night 1.13, 95% CI 0.95–1.34, P = .16).

Unlike differences in risk status due to plurality or birth weight, method of delivery is dependent upon the clinical decision to perform a cesarean delivery. During the early night there was a significant increase in the mortality of infants born vaginally (OR 1.16, 95% CI 1.07–1.26) but not in infants born by cesarean (OR 1.06, 95% CI 0.97–1.17). In the late night, the situation was reversed. Infants delivered by cesarean had a marked increase in mortality (OR 1.32, 95% CI 1.19–1.37, P < .001), whereas those delivered vaginally showed only a weak trend toward increase (OR 1.07, 95% CI 0.99–1.61, P = .098).

Table 3 shows the relationship between time of birth and neonatal mortality by level of neonatal intensive care for hospitals that have official designations. After adjusting for differences in risk across time, there is no significant elevation for infants born in primary care hospitals. In hospitals with intermediate intensive care, infants born in either the early night or the late night have elevated mortality. In hospitals providing community and regional intensive care, mortality is elevated only for infants born in the late night.

Table 4 shows the risk-adjusted relationship between time of birth and death due to specific causes. Deaths reported as being due to asphyxia, not classified, obstetric conditions, prematurity, and other respiratory conditions were increased in both early and late night births. Deaths due to respiratory distress were elevated in the early night. Deaths due to perinatal or neonatal infection were not related to time of birth.

	DISCUSSION

TOP ABSTRACT METHODS RESULTS DISCUSSION REFERENCES

 
Our study demonstrates that increased neonatal mortality in infants born at night is not limited to Europe. After controlling for prenatal care, complications of pregnancy and delivery, gender, and birth weight, our study of California births during 1992–1997 demonstrated a 12% increase in neonatal deaths for infants born during the early night and a 16% increase for infants born during the late night. These excess deaths represent a significant burden because they account for 9.6% of all neonatal deaths in California.

Two possible contributors to excess mortality are fatigue and structural factors. To investigate the role of fatigue, we divided the night into an early "lower fatigue" period and late "higher fatigue" period. If fatigue were an important contributor, we would have expected higher excess mortality for infants born during the late night. Although mortality was elevated for both high- and low-risk infants born at night, the extent of the elevation was similar in early and late night births, suggesting that fatigue did not play a major role (Table 4). This finding is consistent with the literature.⁸ A recent review concludes that, although many studies have documented the adverse consequences of fatigue on personal well-being, self-reported tendency to make medical errors, and performance in clinical simulators, there is a "lack of convincing data linking fatigue with poor patient outcomes."¹² Fatigue does not appear to be a major contributor to the night birth disadvantage based on the pattern of mortality elevation seen in high- and low-risk infants or in deaths due to asphyxia or "other" causes (not classified, obstetric conditions, respiratory distress, other respiratory). However, the 21% mortality increase in hospitals providing community neonatal intensive care and the 14% mortality increase in hospitals providing regional neonatal intensive care were limited to infants born in the late night (Table 3). The extent to which these elevations are a consequence of fatigue or other late night structural factors will require further investigation. Although at the all-California level we are unable to make a strong case for the role of fatigue, it could play an important role in those hospitals where staffing patterns include long shift periods.¹³

The availability and quality of physicians, nurses,¹⁴ and support personnel^15,16 and accessibility of diagnostic tests and procedures^17,18 are examples of structural factors that could adversely affect infants born at night. To investigate "structural competency," we compared the outcomes of lower-demand births (singletons, non-VLBW) with those of higher-demand births (multiple births and VLBW infants). Because multiple births require at least a double resuscitation setup as well as extra personnel to maintain these stations and the birth of a VLBW infant requires personnel with experience in addressing the acute physiologic and pulmonary needs of the premature infant, we hypothesized that, given a decrease in diagnostic and therapeutic capacity, their outcomes would be more severely compromised. Using this paradigm, we found no evidence of a relationship between the extent of required intervention and the extent of mortality increase. To the contrary, the mortality elevation in higher-risk multiples was less extensive then in lower-risk singletons. This paradoxical finding suggests a compromised system that retains some ability to identify and mobilize for a high-risk delivery but fails to optimally address unexpected problems that arise in low-risk deliveries. Although we found no differences in the extent of night birth compromise for VLBW and non-VLBW, the pattern for multiples and singletons is similar to that seen in Sweden where only the lower-risk full term (versus premature) and nonbreech (versus breech) infants had increased mortality when born at night.

Although there is utility in summarizing the effects of hour of birth for the California birth cohort as a whole, there are vast differences in hospitals' structures, staffing, and case mixes, depending upon their designated levels of care. Our analysis demonstrated that time risks differ by neonatal level of care. Primary care hospitals have no licensed neonatal intensive care beds and deliver the least medically complex women. These hospitals have the lowest neonatal mortality rate (1.16), and there was no significant increase in risk-adjusted neonatal mortality rate, suggesting that care for infants born at night was not compromised. Intermediate neonatal intensive care units (NICUs) provide intensive care for neonates who do not require ventilation or surgery and have an overall neonatal mortality rate of 2.09. Infants who require ventilation or complex surgery must be transported to a hospital with a community (ventilation) or regional (ventilation and complex surgery) NICU. After risk adjustment, intermediate care infants born in the early night had a 19% increase in mortality and those born in the late night a 33% increase in mortality. Further research is needed to assess the extent to which this increase may result from nighttime difficulties in arranging and conducting transport to a higher level of care and the extent to which it may be the result of a night case mix that exceeds intermediate care capacity. Community and regional hospitals care for the most complex infants (total birth neonatal mortality rates of 2.20 and 3.85, respectively). In community and regional hospitals, risk-adjusted mortality was only increased in the late night. Although this suggests provider fatigue, structural factors such as the availability of support services may also play a role. We anticipate that a more compete analysis of the differences in risk and specific causes of death across levels of care, although beyond the scope of this initial study, could provide important insights into the factors that increase mortality in night births.

Similar to European studies, we also found that in the cohort as a whole, deaths due to asphyxia were elevated in infants born at night. Extending this observation, we identified that night births had increased deaths due to conditions associated with obstetric care, such as asphyxia, as well as increased deaths in conditions associated with neonatal care, such as respiratory causes and other categories. However, the data set used for our study does not have the information required to assess the extent to which nursing, obstetric, pediatric, anesthesiologic, and diagnostic services may have contributed to the night birth disadvantage.

Although we did not identify strong evidence for a late night fatigue effect, there were temporal differences in outcomes by mode of delivery. In the early evening, cesareans decreased from 24.4% during the day to 19.8%. Case mix–adjusted mortality was increased, but only in infants delivered vaginally, perhaps suggesting that some infants delivered vaginally may have benefited from cesarean delivery. During the late evening, only 12.7% of deliveries were by cesarean. Unexpectedly, even with this very dramatic decrease in the cesarean rate (in part due to scheduling elective cesareans in the daytime and early evening), there was only a trend toward an increase in the mortality of vaginally delivered infants (OR 1.07, 95% CI 0.99–1.61, P = .098). However, infants delivered by cesarean had a 32% increase in mortality (OR 1.32, 95% CI 1.19–1.47). Although the lack of an elevation in the mortality of vaginally delivered infants argues that the majority of those infants who would benefit by cesarean had been selected for cesarean delivery, the 32% mortality increase in late night births by cesarean suggests that there may have been a delay in the timing and/or compromised conduct of those appropriately selected.¹⁹ Further research will be required to understand these complex findings.

Our study has several limitations. Because there is no universal nursing, support staff, or physician shift pattern in California, our designations for day, early night, and late night were arbitrarily decided at the inception of the study. Although they may not provide an optimal fit, these periods are in keeping with the neonatal mortality rate time trends seen in Figure 1. Our analyses addressed only the time of day and did not consider possible weekend, holiday, or seasonal effects. A second limitation is that our data set did not allow us to examine intrapartum deaths, a sensitive indicator of the quality of perinatal care.²⁰ Although Swedish intrapartum deaths were not elevated at night,¹ we believe that the extent of the night birth mortality disadvantage and its specific expression as elevated intrapartum, early neonatal, and/or neonatal mortality will be dependent upon both health care system and specific hospital factors. Although the findings from an individual country, state, or health care system serve to draw attention to an important area of concern, it is neither possible nor prudent to generalize these findings across health care systems. It will, therefore, be important to assess the degree to which the night birth disadvantage described for European and Californian births occurs in other areas of the United States. A third limitation is that our data did not permit an evaluation of the sources of this disadvantage. Further research is needed to identify the specific differences in structure and process that result in the night birth disadvantage. Finally, although we attempted to adjust for differences in case mix by using risk-stratified (VLBW versus non-VLBW, singleton versus multiple, vaginal versus cesarean delivery) logistic regressions that controlled for temporal differences in birth weight, gender, prenatal care, and perinatal complications obtained from the birth certificate, it is possible that some portion of the night birth disadvantage may be due to risk features that we were unable to identify from our data set. It is also possible that some portion may be due to underreporting of the risk factors that we controlled for. Although studies from Washington state²¹ and New York²² concluded that complications of pregnancy were underreported, there have been no California validation studies for these items.

Even with these caveats, our study provides strong evidence that infants born at night have at 12–16% increase in mortality. Because this disadvantage accounts for almost 10% of all neonatal deaths in California, assessing the extent to which infants born at night contribute to the national neonatal mortality rate and addressing this temporal disparity are high priorities for perinatal medicine. From a more immediate clinical perspective, we recommend that perinatal morbidity and mortality case conferences consider time of birth as an important area of inquiry in their investigation of contributory factors.

	Footnotes

 
We thank the Maternal and Child Health Branch, California Department of Health Services, for their support of the Perinatal Profiles project (contract 99-85027).

Address reprint requests to: Jeffrey B. Gould, MD, MPH, Division of Neonatal and Developmental Medicine, Stanford University, 750 Welch Road, Suite 315, Palo Alto, CA 94304; jbgould{at}stanford.edu.

doi:10.1097/01.AOG.0000168627.33566.3c

	REFERENCES

TOP ABSTRACT METHODS RESULTS DISCUSSION REFERENCES

 
1. Stephansson O, Dickman PW, Johansson AL, Kieler H, Cnattingius S. Time of birth and risk of intrapartum and early neonatal death. Epidemiology 2003;14:218–22.[Medline]

2. Luo ZC, Karlberg J. Timing of birth and infant and early neonatal mortality in Sweden 1973–95: longitudinal birth register study. BMJ 2001;323:1327–30.[Abstract/Free Full Text]

3. Heller G, Misselwitz B, Schmidt S. Early neonatal mortality, asphyxia related deaths, and timing of low risk births in Hesse, Germany, 1990–8: observational study. BMJ 2000;321:274–5.[Free Full Text]

4. Stewart JH, Andrews J, Cartlidge PH. Numbers of deaths related to intrapartum asphyxia and timing of birth in all Wales perinatal survey, 1993–5. BMJ 1998;316:657–60.[Abstract/Free Full Text]

5. Chalmers JW, Shanks E, Paterson S, McInneny K, Baird D, Penney G. Scottish data on intrapartum related deaths are in same direction as Welsh data. BMJ 1998;317:539–40.[Free Full Text]

6. Paccaud F, Martin-Beran B, Gutzwiller F. Hour of birth as a prognostic factor for perinatal death. Lancet 1988;1:340–3.[Medline]

7. Chamberlain G. The night: a dangerous time to be born? Br J Obstet Gynaecol 1987;94:922–4.[Medline]

8. Gaba DM, Howard SK. Patient safety: fatigue among clinicians and the safety of patients. N Engl J Med 2002;347:1249–55.[Free Full Text]

9. Tashiro M. Tape documentation of perinatal record format California Birth Cohort, 1996. Sacramento, CA: California Department of Health Services, Maternal and Child Health Branch; 1999.

10. California Childrens Services. CCS approved neonatal intensive care units. Sacramento, CA: Department of Health Services, Childrens Medical Services, California Childrens Services; 2000. p. 1–29.

11. Dollfus C, Patetta M, Siegel E, Cross AW. Infant mortality: a practical approach to the analysis of the leading causes of death and risk factors. Pediatrics 1990;86:176–83.[Abstract/Free Full Text]

12. Jha AK, Duncan BW, Bates DW. Fatigue, sleepiness, and medical errors. In: Shojania KG, McDonald KM, Wachter RM, ed. Making health care safer: a critical anaysis of patient safety practices. Rockville, Md: Agency for Healthcare Research and Quality; 2001. p. 519–31.

13. Landrigan CP, Rothschild JM, Cronin JW, Kaushal R, Burdick E, Katz JT, et al. Effect of reducing interns' work hours on serious medical errors in intensive care units. N Engl J Med 2004;351:1838–48.[Abstract/Free Full Text]

14. Sochalski J. Is more better? The relationship between nurse staffing and the quality of nursing care in hospitals. Med Care 2004;42:II67–73.[Medline]

15. Dexter F, Epstein RH, Marsh HM. Costs and risks of weekend anesthesia staffing at 6 independently managed surgical suites. AANA J 2002;70:377–81.[Medline]

16. Bond CA, Raehl CL, Franke T. Clinical pharmacy services, hospital pharmacy staffing, and medication errors in United States hospitals. Pharmacotherapy 2002;22:134–47.[Medline]

17. Ward-Cook K. Medical laboratory workforce trends and projections: what is past is prologue. Clin Leadersh Manag Rev 2002;16:364–9.[Medline]

18. Boyd JC, Savory J. Genetic algorithm for scheduling of laboratory personnel. Clin Chem 2001;47:118–23.[Abstract/Free Full Text]

19. Gould JB, Danielsen B, Korst LM, Phibbs R, Chance K, Main E, et al. Cesarean delivery rates and neonatal morbidity in a low-risk population. Obstet Gynecol 2004;104:11–9.[Abstract/Free Full Text]

20. Kiely JL, Paneth N, Susser M. Fetal death during labor: an epidemiologic indicator of level of obstetric care. Am J Obstet Gynecol 1985;153:721–7.[Medline]

21. Dobie SA, Baldwin LM, Rosenblatt RA, Fordyce MA, Andrilla CH, Hart LG. How well do birth certificates describe the pregnancies they report? The Washington State experience with low-risk pregnancies. Matern Child Health J 1998;2: 145–54.[Medline]

22. Roohan PJ, Josberger RE, Acar J, Dabir P, Feder HM, Gagliano PJ. Validation of birth certificate data in New York State. J Community Health 2003;28:335–46.[Medline]

This article has been cited by other articles:

				L. Rautava, L. Lehtonen, M. Peltola, E. Korvenranta, H. Korvenranta, M. Linna, M. Hallman, S. Andersson, M. Gissler, J. Leipala, et al. The Effect of Birth in Secondary- or Tertiary-Level Hospitals in Finland on Mortality in Very Preterm Infants: A Birth-Register Study Pediatrics, January 1, 2007; 119(1): e257 - e263. [Abstract] [Full Text] [PDF]

				J. M. Dodd, C. A. Crowther, and J. S. Robinson Morning Compared With Evening Induction of Labor: A Nested Randomized Controlled Trial Obstet. Gynecol., August 1, 2006; 108(2): 350 - 360. [Abstract] [Full Text] [PDF]

				Y. W. Wu, L. A. Croen, S. J. Shah, T. B. Newman, and D. V. Najjar Cerebral Palsy in a Term Population: Risk Factors and Neuroimaging Findings Pediatrics, August 1, 2006; 118(2): 690 - 697. [Abstract] [Full Text] [PDF]

				M. E. Abdel-Latif, B. Bajuk, J. Oei, K. Lui, and for the New South Wales and the Australian Capital Mortality and Morbidities Among Very Premature Infants Admitted After Hours in an Australian Neonatal Intensive Care Unit Network Pediatrics, May 1, 2006; 117(5): 1632 - 1639. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Gould, J. B. Articles by Chavez, G. Search for Related Content PubMed PubMed Citation Articles by Gould, J. B. Articles by Chavez, G.