Morphine Administration and Short-term Pulmonary Outcomes Among Ventilated Preterm Infants
http://www.100md.com
《小儿科》
Department of Pediatrics, Albert Einstein Medical Center, Philadelphia, Pennsylvania
Karolinska Institute, Neonatal Research Unit, Astrid Lindgren's Children's Hospital, Stockholm, Sweden
Maryland Medical Research Institute, Baltimore, Maryland
Department of Pediatrics, University of Arkansas for Medical Sciences, Arkansas Children's Hospital, Little Rock, Arkansas
ABSTRACT
Background. The use of opioid therapy for sedation and analgesia among ventilated infants varies among care providers. The impact of opioid therapy early in the neonatal course of respiratory distress syndrome (RDS) on pulmonary outcomes is not known.
Objective. We tested the hypothesis that preterm neonates randomized to the morphine infusion group would have improved ventilatory outcomes, measured as shorter durations of ventilator or oxygen therapy, fewer air leaks, and lower incidence of bronchopulmonary dysplasia.
Methods. All 898 subjects (gestational age [GA] of 23 to 32 weeks) who were enrolled in the Neurologic Outcomes and Preemptive Analgesia in Neonates (NEOPAIN) trial formed the study cohort (morphine: 449 patients; placebo: 449 patients). Subjects received the masked study drug until they were weaned from the ventilator or for 14 days, whichever occurred earlier. Outcome measures included air leaks, duration of ventilation or oxygen therapy, hospitalization, bronchopulmonary dysplasia, and death.
Results. Subjects in the 2 groups had similar baseline characteristics (mean ± SD, morphine versus placebo: GA: 27.3 ± 2.3 vs 27.4 ± 2.3 weeks; birth weight: 1037 ± 340 vs 1054 ± 354 g). Infants in the morphine group required ventilator therapy significantly longer, compared with the placebo group (median [interquartile range]: 7 days [4–20 days] vs 6 days [3–19 days]). This difference in ventilation duration was significant for infants with GA of 27 to 29 weeks (6 days [4–12 days] vs 5 days [2–9 days]) and 30 to 32 weeks (4 days [3–6 days] vs 3 days [2–5 days]). Infants who received additional analgesia with intermittent morphine doses in both groups were sicker than those who were not given open-label morphine. After adjustment for birth weight, Clinical Risk Index for Babies scores, maternal chorioamnionitis, RDS requiring surfactant, and patent ductus arteriosus in a logistic regression model, the use of additional analgesia with morphine was associated independently with increased air leaks and longer durations of high-frequency ventilation, nasal continuous positive airway pressure, and oxygen therapy.
Conclusions. Morphine infusions do not improve short-term pulmonary outcomes among ventilated preterm neonates. Additional morphine doses were associated with worsening respiratory outcomes among preterm neonates with RDS.
Key Words: pain bronchopulmonary dysplasia lung newborn infant morphine
Abbreviations: AA, additional analgesia BPD, bronchopulmonary dysplasia BW, birth weight CPAP, continuous positive airway pressure CRIB, Clinical Risk Index for Babies GA, gestational age HFV, high-frequency ventilation M3G, morphine-3-glucuronide M6G, morphine-6-glucuronide nCPAP, nasal continuous positive airway pressure NEOPAIN, Neurologic Outcomes and Preemptive Analgesia in Neonates NOPAIN, Neonatal Outcome and Prolonged Analgesia in Neonates PDA, patent ductus arteriosus RDS, respiratory distress syndrome
The relief of pain per se is certainly an appropriate clinical goal for premature neonates.1 There is, however, significant variation in the use of opioid analgesics in NICUs.2 The impact of opioid therapy, used early in the neonatal course, on short-term outcomes is not known.3 The studies4–7 that have evaluated the impact of morphine on neonatal ventilatory outcomes have varied in their findings. Dyke at al4 reported that morphine therapy was associated with improved ventilator synchrony and reduced heart rate, respiratory rate, and duration of oxygen therapy among ventilated preterm neonates. That study, however, had only 26 infants enrolled and the results of that small pilot study should be interpreted with caution.4 Our pilot study, the Neonatal Outcome and Prolonged Analgesia in Neonates (NOPAIN) trial, suggested improved neurologic outcomes for the group of patients who received morphine (vs midazolam or placebo) but no differences in the number of days required for mechanical ventilation, continuous positive airway pressure (CPAP), or oxygen therapy.5
Among 20 ventilated premature infants randomized to receive fentanyl or placebo infusions, those who received fentanyl required a longer duration of mechanical ventilation but showed no effects on bronchopulmonary dysplasia (BPD), intraventricular hemorrhage, or sepsis.6 In another study of 27 infants (compared with 28 untreated control subjects), there were no significant differences in the short-term ventilatory parameters between the 2 groups.7 However, these studies4–7 had small numbers of subjects and inadequate power to detect or to reject any difference in these outcomes. Therefore, there is no reliable evidence regarding the effects of morphine on pulmonary outcomes among neonates.
The subjects enrolled in the Neurologic Outcomes and Preemptive Analgesia in Neonates (NEOPAIN) multicenter trial provided the opportunity to explore the effects of opioid therapy on pulmonary outcomes. We proposed, a priori, to test the hypothesis that preterm infants with respiratory distress syndrome (RDS) who received morphine infusions would have shorter durations of ventilation, CPAP, or oxygen therapy and decreased rates of pulmonary air leaks or BPD.
METHODS
Study Group
Methodologic details of the NEOPAIN trial were published recently.8 Briefly, preterm neonates who were born between 23 and 32 weeks of gestation, intubated before 72 hours of age, and ventilated for <8 hours at the time of enrollment were eligible. Neonates with major congenital anomalies, birth asphyxia, intrauterine growth restriction (<5th percentile), or maternal opioid addiction and those participating in other clinical trials were excluded. Written parental consent was obtained for all study infants. The NEOPAIN protocol and consent form were approved by local ethics committees at each of the 16 participating NICUs and by an external ethics committee at the coordinating center. An independent data and safety monitoring board was constituted to monitor the trial.
Randomization was performed with an automated telephone response system and was stratified according to the participating NICU and gestational age (GA) (23–26 weeks, 27–29 weeks, or 30–32 weeks), to ensure equal numbers in the morphine and placebo groups. To eliminate clinical bias, all clinical personnel were blinded to the study drug code, and unblinding of the treatment code (n = 6) was limited by stringent criteria.
Therapeutic Management
Data Collection
Data collection included baseline clinical and demographic characteristics (Table 1), vital signs, and clinical outcomes related to ventilator management. The clinical factors defined below were selected a priori, on the basis of clinically reported associations with mechanical ventilation or BPD.
Definitions
Maternal Variables
A positive culture from the amniotic fluid was defined as chorioamnionitis. Chorioamnionitis was also diagnosed on the basis of positive Gram stain results, positive white blood cell counts (50 cells per mm3), or low glucose concentrations (14 mg/dL) in the amniotic fluid. Chorioamnionitis was suspected if any 2 of the following clinical symptoms were present: maternal fever, leukocytosis, uterine tenderness, foul-smelling amniotic fluid, or fetal tachycardia.
Neonatal Variables
RDS was diagnosed on the basis of the clinical findings of respiratory distress and chest radiographs showing a diffuse reticulogranular pattern with air bronchograms. The Clinical Risk Index for Babies (CRIB), which includes birth weight (BW), GA, congenital malformations, maximal base excess, and minimal and maximal inspired oxygen concentrations in the first 12 hours, was calculated.17 Durations of mechanical ventilation and/or nasal CPAP (nCPAP) were defined as the completed days of these therapies. The number of doses of surfactant given for the treatment of RDS served as a surrogate measure of the severity of RDS. Patent ductus arteriosus (PDA) was diagnosed on the basis of clinical signs of high-output congestive heart failure and was confirmed with echocardiographic evidence of ductal patency. Air leaks (pneumothorax, pneumomediastinum, pulmonary interstitial emphysema, pneumoperitoneum, pneumopericardium, or subcutaneous emphysema) were diagnosed on the basis of chest radiographic findings, ultrasonographic findings, or clinical signs. Early-onset sepsis was defined as a positive blood culture in the first 72 hours with clinical signs of sepsis or strong clinical evidence of sepsis with abnormal hematologic values but no positive culture. BPD was diagnosed on the basis of clinical and chest radiographic findings of BPD and the need for supplemental oxygen at a corrected GA of 36 weeks. Length of stay was defined as the number of completed days in the hospital.
Statistical Analyses
The hypotheses, study protocol, sample size, and methods for this prospective study were determined a priori before the initiation of statistical analyses. Assuming an incidence of BPD in the target population (GA of 23–32 weeks) of 30%, we estimated that 370 infants in each group would allow us to detect a change of at least 30% at = .05 and a power of 0.8. Treatment group demographic features and outcomes were compared with Student's t test, 2 tests, or, for contingency tables with small cell numbers, Fisher's exact tests. For data that did not have a strictly normal distribution, the results are presented as median and interquartile range, and nonparametric tests (Wilcoxon test and Kruskal-Wallis test) were used for analyses (Tables 1–4), as appropriate.
Adjusting for BW and CRIB scores, we analyzed the effects of AA (intermittent morphine boluses) and clinical factors (maternal chorioamnionitis, RDS requiring surfactant, and PDA) on the days of ventilation, nCPAP, and oxygen treatment, as well as the length of stay, with multiple linear regression models (Table 5). Each predictor factor was entered into the models as a binary (0 or 1) variable, with 1 indicating the presence of the factor and 0 indicating the absence of the factor. The CRIB score, with a range of 0 to 20, was entered into the models as a ranking variable, with 10 as the reference score. Results of the multiple linear regression models are presented as parameter estimates and SEs.
Factors associated with the outcomes of pulmonary air leaks, BPD, and BPD or death were analyzed with logistic models (Table 5). The fit of each logistic model was assessed with the Hosmer-Lemeshow goodness-of-fit test, and the global test that all regression parameters were 0 was tested with the –2 log-likelihood statistic. Results of these analyses are presented as odds ratios with 2-sided 95% confidence intervals, to show the effect of each predictor variable on the indicated outcome. Results are presented for covariates that were statistically significant. All analyses were performed with SAS statistical software (SAS Institute, Cary, NC), with the critical P value set at .05.
RESULTS
Data on 898 infants were available for analysis, with 449 in the morphine group and 449 in the placebo group. Demographic data are shown in Table 1. The 2 groups were well matched for maternal and infant factors (Table 1).
The neonatal outcomes of all infants are presented in Table 2. The frequency of RDS and the overall severity of illness (CRIB scores) were similar in the 2 groups. Significantly more days were spent on mechanical ventilation in the morphine group, compared with the placebo group (P < .01), as reported earlier.8 No differences in the durations of nCPAP, oxygen therapy, or hospitalization or in the incidences of air leaks, early-onset sepsis, total sepsis, postnatal steroid use, BPD, or deaths occurred between the 2 randomized groups (Table 2).
In the 3 different GA strata (23–26 weeks, 27–29 weeks, and 30–32 weeks), there were no differences in the incidence of RDS, number of surfactant doses, incidence of PDA, indomethacin use, incidence of early-onset or total sepsis, length of stay, or number of deaths between the morphine and placebo groups (data not shown). In the 23 to 26 weeks' GA category, there were no differences between the 2 groups in the total ventilation days (median [interquartile range]: 18 days [7–39 days] vs 20 days [6–37.5 days]; P = .94) or duration of oxygen use (55 days [11–83 days] vs 55 days [18–78 days]; P = .69). The incidences of air leaks (21% vs 19%) and BPD (38% vs 36%) were not significantly different in the morphine and placebo arms.
In the 27 to 29 weeks' GA category, the morphine group required longer ventilation than did the placebo group (6 days [4–12 days] vs 5 days [2–9 days]; P < .01), but no difference was noted in the duration of oxygen use (31 days [13–52 days] vs 34 days [14–49.5 days]; P = .95). The incidences of air leaks (7% vs 6%) and BPD (17% vs 18%) were not significantly different in the morphine and placebo arms. In the 30 to 32 weeks' GA category, the morphine group required longer ventilation than did the placebo group (4 days [3–6 days] vs 3 days [2–5 days]; P = .02), but no difference was noted in the duration of oxygen use (8 days [4–25 days] vs 7.5 days [3–15 days]; P = .25). The incidences of air leaks (5% vs 10%) and BPD (6% vs 7%) were not significantly different in the morphine and placebo arms. Differences in the duration of ventilation remained significant for the 27 to 29 weeks' GA (6 days [4–12 days] vs 5 days [2–9 days]; P < .01) and 30 to 32 weeks' GA (4 days [3–6 days] vs 3 days [2–5 days]; P = .03) strata, even after exclusion of deaths from both randomized groups.
To evaluate the effect of open-label AA in the morphine and placebo groups, data were analyzed in each group by comparing infants who did versus did not receive AA. Their demographic data are noted in Table 3. Infants who received AA, in both randomized groups, had significantly lower maternal age, less maternal chorioamnionitis, lower GA and BW, and lower Apgar scores at 1 and 5 minutes (Table 3).
A higher incidence of RDS and greater severity of illness (CRIB scores) were noted for infants who received AA (Table 4). Therefore, infants who received AA in either randomized group were "sicker" and had greater physiologic instability. This might have led to increased complications with more frequent procedures, resulting in increased assessments of pain by their caregivers and consequently increased use of AA. Table 4 shows significantly higher incidences of PDA, air leaks, and chest tubes and worse respiratory outcomes, with longer durations of ventilation, oxygen therapy, and hospital stay, higher incidence of BPD, and more neonatal deaths, among infants who received AA, compared with those who did not.
To examine whether morphine therapy contributed to the respiratory outcomes outlined above, we adjusted for BW, CRIB scores (which include GA in the composite scores), maternal chorioamnionitis, RDS requiring surfactant, and PDA in our logistic regression models, using placebo/no AA as the reference group. Preemptive morphine infusions did not alter these outcomes, but the use of morphine AA continued to be a strong independent predictor of worse respiratory outcomes even after adjustment for the aforementioned factors (Table 5).
Infants who received AA in the placebo group had a 3.4-fold increase in pulmonary air leaks (P < .01), compared with infants who received placebo and no AA. On average, these infants spent an additional 2.6 days on HFV (P < .01) and 3.2 days on nCPAP (P = .02) and required supplemental oxygen for an additional 7.2 days (P = .01), compared with infants who received placebo and no AA (Table 5).
Infants who received AA in the morphine group had even worse respiratory outcomes (Table 5), with a 4.3-fold increase in air leaks (P < .01), compared with infants who received placebo and no AA. On average, infants who received AA in the morphine group spent an additional 6.7 days on positive-pressure ventilation (P < .01), with similar contributions from conventional ventilation (P = .04) and HFV (P < .01). These infants also spent more time on nCPAP (additional 2.9 days; P = .04) and on supplemental oxygen (additional 8.1 days; P < .01).
DISCUSSION
A previous study suggested that morphine infusions increase the synchronicity of spontaneous and ventilator-derived breaths among preterm infants with RDS and reduce the duration of oxygen therapy.4 In such a scenario, if morphine analgesia decreases the ventilatory pressures and need for oxygen, then potentially it could reduce ventilator-induced lung injury, perhaps resulting in a decreased incidence of BPD.
Results from the NEOPAIN multicenter trial suggest that use of preemptive morphine analgesia among premature infants with RDS does not decrease the incidences of pulmonary air leaks, BPD, or other respiratory outcomes. In a randomized, double-blind trial of morphine versus diamorphine among ventilated preterm neonates (n = 88), the 2 drug regimens were equally effective in terms of sedation, with no significant differences in mortality rates, ventilator days, or BPD rates.18 In the NOPAIN pilot study, no significant differences in ventilatory outcomes among the morphine (n = 24), midazolam (n = 22), and placebo (n = 21) groups were noted, but the durations of ventilation, nCPAP, and oxygen therapy were lowest in the morphine group.5
Ventilatory indices were unchanged in a short-term study comparing ventilated infants with RDS who received midazolam (n = 24) versus placebo (n = 22).19 Among mechanically ventilated neonates with RDS who did not receive surfactant, infants who received morphine (n = 11) had improved outcomes, compared with those who received chloral hydrate (n = 8), as evidenced by lower mean airway pressures, ventilator rates, and need for oxygen and improved survival rates in the morphine group.20
In our study, overall there was no difference in the incidence of BPD, despite significant differences in the duration of ventilation (Table 2). This highlights the complex and multifactorial nature of BPD.21 It is noteworthy that the incidence of BPD was higher among the infants who received more morphine (Table 4). Interestingly, after correction for variables known to affect BPD, AA with morphine continued to be associated significantly with longer duration of ventilation and oxygen therapy but not with the incidence of BPD (Table 5). Therefore, traditional environmental factors (for example, ventilator-induced lung injury, hyperoxia, sepsis, and PDA) may not be solely responsible and may require association with other (for example, genetic) influences to cause the disordered lung development that is the hallmark of BPD.21
On the basis of currently available physiologic/behavioral markers, opioid analgesia seems definitely to reduce pain/stress among ventilated preterm neonates.5–7,18 In terms of pulmonary outcomes, previous data suggested that opioid analgesia may be preferable over sedation,5,20 but the combination of opioid analgesia and muscle relaxation and/or sedative agents should be used with caution among ventilated neonates.22,23 Furthermore, a meta-analysis has raised concerns regarding the use of midazolam infusions among neonates.24
Most studies examining respiratory outcomes among ventilated neonates receiving analgesia enrolled a small number of infants5–7,19,20 and/or lacked a placebo arm.3,18,20 A recent randomized trial of morphine (n = 77) versus placebo (n = 73) among premature ventilated infants did not show any differences in the duration of ventilation or the incidence of BPD between the 2 groups.25 The large cohort of infants in the NEOPAIN trial (with a placebo arm) provided a more definitive opportunity to assess the effects of opioid analgesia on the respiratory outcomes of ventilated preterm infants.
Infants receiving morphine infusions were ventilated for significantly longer periods than were those receiving placebo, with differences primarily among 27- to 29-week neonates (receiving 20 μg/kg per hour morphine) and 30- to 32-week neonates (receiving 30 μg/kg per hour morphine). This most likely reflects the respiratory depression caused by increased doses of morphine, compared with that for 23- to 26-week neonates, who received 10 μg/kg per hour morphine. It is possible that the doses of morphine administered in the present study were excessive for the population studied. This could be a result of the dosing regimen or drug metabolism (see below). Weaning of the study drug was performed only for factors such as extreme suppression of the respiratory drive, and it is quite likely that the effects of the morphine infusion included not only analgesia but also excessive sedation. The results indicating a longer duration of mechanical ventilation in the morphine group would then be logical, because infants in the morphine group would be more sedated and therefore would have a lower respiratory drive.
The use of infant-triggered respiratory rate as a measure of respiratory drive is a novel approach to studying the problem of respiratory depression with morphine, although it is dependent on the infant's interaction with the ventilator and the type of triggering devices in the ventilator.26 Among trigger-ventilated preterm neonates (n = 14; BW: 1.37 kg; GA: 30 weeks) who received morphine (100 μg/kg bolus and 10 μg/kg/hour infusion), infants who had morphine-6-glucuronide (M6G) detectable at 12 hours (n = 6) showed a significantly greater reduction in triggered breath rate (–22 beats per minute), compared with those who did not (n = 3; –4 beats per minute) (P = .03).26 This raises questions about the role of metabolites of morphine in producing its clinical effects. Morphine is metabolized in the liver to the polar conjugates morphine-3-glucuronide (M3G) and M6G. The formation of the glucuronides is slow in preterm newborns because of immature liver function, but their half-lives are prolonged because of poor renal excretion. Despite their poor lipophilicity, there is clear evidence of brain penetration by the glucuronides,27,28 and this is likely to occur among preterm newborns with a poorly developed blood-brain barrier. There is evidence that M6G is a more potent analgesic agent and respiratory depressant than morphine itself,29,30 whereas M3G has antiopioid effects, although this evidence is not consistent.31,32 The formation of M3G in preterm newborns occurs earlier and in greater amounts than that of M6G.26 Therefore, in assessments of the respiratory depressant effects of morphine among preterm neonates, variables such as hepatic metabolism and renal excretion may be more important than the dosing regimen.
In a recent double-blind study comparing continuous infusion of 10 μg/kg per hour versus intermittent morphine boluses of 30 μg/kg every 3 hours among term neonates (and older children), it was reported that neonates had a narrower therapeutic window for postoperative morphine analgesia than did older age groups, with no difference in the safety or effectiveness of intermittent doses, compared with continuous infusions.33 The high plasma concentrations of morphine metabolites in the neonates were the result of low renal clearance, which was confirmed by the significant correlation between serum creatinine and M6G levels in this age group. The decreased clearance of morphine and M6G explained the increased analgesic effects among neonates.33 The authors concluded that morphine given intermittently does not provide any clinical advantages and that a continuous morphine infusion is probably safer for neonates.33 Another report from the same investigators described the effects of age and other clinical factors among 68 neonates (52 who were <7 days of age and 16 who were 7 days of age) after major surgery.34 Although pain scores were no different, the younger neonates had different morphine requirements (10 vs 10.8 μg/kg per hour), plasma morphine concentrations (23.0 vs 15.3 ng/mL), and M6G/morphine ratios (0.6 vs 1.5).34 Respiratory insufficiency occurred for 5 neonates who received intermittent morphine, but this difference between the groups was not significant. The authors concluded that neonates 7 days of age required significantly less morphine postoperatively than older neonates.34
These data suggest that patients of younger GA and sicker infants (requiring mechanical ventilation) would have lower requirements for morphine and that intermittent boluses of morphine might have more adverse effects. It is noteworthy that the infants of youngest GA (23–26 weeks) in our study were receiving 10 μg/kg per hour, the same dose as the term neonates in the aforementioned 2 reports.33,34
Another important result, apart from prolonged ventilation associated with a higher dose of morphine infusion, was the fact that morphine AA was an independent predictor of worse neonatal respiratory outcomes, even after controlling for the effects of BW and severity of illness (CRIB score). The CRIB score was superior to BW35 or the Score for Neonatal Acute Physiology for predicting survival to discharge.36
Interestingly, logistic regression analyses showed that the respiratory outcomes of infants who received preemptive morphine infusions alone (no AA) were not significantly different from those of the placebo/no AA group. In contrast, infants who received intermittent doses of morphine (AA) had worse respiratory outcomes, regardless of whether they received placebo or morphine infusions (Table 5). This effect was somewhat accentuated in the morphine group with AA, which suggests a cumulative dose effect of morphine on neonatal respiratory outcomes. This association of AA with worse pulmonary outcomes does not imply causation, because other unknown and/or unrecorded factors might be contributing to the results.
Because morphine AA was given at the discretion of the caregiver (based on their evaluation of "pain/discomfort" in the neonate), our data suggest that, for extremely sick, ventilated, premature newborns, providing "pain relief" through intermittent boluses of morphine might worsen their outcomes. We cannot discount completely the contribution of disease severity in this population to worse respiratory outcomes, but intermittent boluses of morphine seem to worsen neonatal ventilatory outcomes independently. We suggest caution in the use of intermittent morphine boluses for treating "pain/discomfort" in such populations.
CONCLUSIONS
Our data suggest that use of morphine analgesia for ventilated preterm infants does not reduce the incidence of BPD or other respiratory outcomes. Morphine infusions of >10 μg/kg per hour might prolong the period of ventilation for preterm neonates, compared with control subjects. Furthermore, intermittent boluses of morphine were associated with worsening respiratory outcomes among ventilated preterm neonates.
ACKNOWLEDGMENTS
This study was supported by grants from the National Institute for Child Health and Human Development (HD36484 to K.J.S.A. and HD36270 to B.A.B.), from the Chief Scientist's Office of the Scottish Executive (to N. McIntosh), from the Vardal Foundation, Free Masons, and Swedish Medical Association, Sweden (to H. Lagercrantz and L.L.B.), from the Fondation pour la Santé CNP, France (to R. Carbajal and R. Lenclen), and from the rebro University Hospital Research Foundation, Sweden (to J. Schollin and M. Eriksson).
Centers and participants in the NEOPAIN trial were: United States: Arkansas Children's Hospital (Little Rock, AR), K.J.S. Anand, J. Seibert, M.B. Moore, H. Farrar, J. Valentine, T.J. Green, L. Letzig; Maryland Medical Research Institute (Baltimore, MD), B.A. Barton, S.S. Kronsberg, S. Fick; University of Arkansas for Medical Sciences (Little Rock, AR), R.W. Hall, R. Arrington, R. Smith; University of Kentucky (Lexington, KY), N. Desai, V.L. Whitehead, J. L. Sampers, P.E. DeFranco, L.A. Shook, T.H. Pauly; Tufts University and New England Floating Hospital, B.A. Shephard, B. Mackinnon, J. Fiascone; Wake Medical Center (Raleigh, NC), T.E. Young, K. Carr, M.R. Johnson, K. McDaniel; Mercy Hospital and Medical Center (Chicago, IL), R. Vasa, J.S. Teji, D. Jahn; University of Pennsylvania and Pennsylvania Hospital (Philadelphia, PA), V.K. Bhutani, J. R. Gerdes, S. Abbasi, M.K.S. Grous, A. Scwhoebel; Medical Center of Delaware, D.J. Tuttle, K.H. Leef, J.L. Stefano; University of Illinois at Chicago (Chicago, IL), R. Bhat, T.N.K. Raju, G. Chari; Albert Einstein Medical Center (Philadelphia, PA), V. Bhandari, H. Hurt, J. Giannetta; Brigham and Women's Hospital (Boston, MA), E.C. Eichenwald, K. Housefield; Long Beach Memorial Hospital (Long Beach, CA), G. Padilla, K. Norris; France: Centre Hospitalier Poissy Saint Germain, Site Poissy, R. Carbajal, R. Lenclen, M. Jugie; Sweden: Karolinska Institute at Astrid Lindgren's Children's Hospital (Stockholm, Sweden), H. Lagercrantz, L.L. Bergqvist; rebro University Hospital (rebro, Sweden), J. Schollin, M. Eriksson; United Kingdom: Simpson Memorial Maternity Pavilion (Edinburgh, Scotland), N. McIntosh, G. Menon, E.M. Boyle; Data and Safety Monitoring Board: R. Troug (Harvard Medical School and Children's Hospital, Boston, MA), A. Aynsley-Green (Institute of Child Health and Hospital for Sick Children, London, United Kingdom), K.D. Craig (University of British Columbia, Vancouver, Canada), C.C. Johnston (McGill University School of Nursing and Montreal Children's Hospital, Montreal, Canada), M. Walden (Baylor University and Texas Children's Hospital, Houston, TX); W.A. Silverman (Greenbrae, CA); E. Bancalari (University of Miami, FL); J. Rowe (National Institutes of Health, Bethesda, MD, and Office on Women's Health, Washington DC).
We thank all of the physicians, nurses, pharmacists, ultrasonographers, occupational therapists, and physical therapists at the participating institutions and the parents who gave consent for this study.
FOOTNOTES
Accepted Nov 24, 2004.
No conflict of interest declared.
REFERENCES
Anand KJS. Clinical importance of pain and stress in preterm neonates. Biol Neonate. 1998;73 :1 –9
Kahn DJ, Richardson DK, Gray JE, et al. Variation among neonatal intensive care units in narcotic administration. Arch Pediatr Adolesc Med. 1998;152 :844 –851
Kennedy KA, Tyson JE. Narcotic analgesia for ventilated newborns: are placebo-controlled trials ethical and necessary J Pediatr. 1999;134 :127 –129
Dyke MP, Kohan R, Evans S. Morphine increases synchronous ventilation in preterm infants. J Paediatr Child Health. 1995;31 :176 –179
Anand KJS, McIntosh N, Lagercrantz H, et al. Analgesia and sedation in preterm neonates who require ventilatory support: results from the NOPAIN trial. Arch Pediatr Adolesc Med. 1999;153 :331 –338
Orsini AJ, Leef KH, Costarino A, Dettore MD, Stefano JL. Routine use of fentanyl infusions for pain and stress reduction in infants with respiratory distress syndrome. J Pediatr. 1996;129 :140 –145
Lago P, Benini F, Agosto C, Zacchello F. Randomised controlled trial of low dose fentanyl infusion in preterm infants with hyaline membrane disease. Arch Dis Child Fetal Neonatal Ed. 1998;79 :F194 –F197
Anand KJS, Hall RW, Desai NS, et al. Effect of morphine analgesia in ventilated preterm neonates: primary outcomes from the NEOPAIN randomized trial. Lancet. 2004;263 :1673 –1682
Bhat R, Abu-Harb M, Chari G, Gulati A. Morphine metabolism in acutely ill preterm newborn infants. J Pediatr. 1992;120 :795 –799
Barrett DA, Barker DP, Rutter N, Pawula M, Shaw PN. Morphine, morphine-6-glucuronide and morphine-3-glucuronide pharmacokinetics in newborn infants receiving diamorphine infusions. Br J Clin Pharmacol. 1996;41 :531 –537
Barrett DA, Elias-Jones AC, Rutter N, Shaw PN, Davis SS. Morphine kinetics after diamorphine infusion in premature neonates. Br J Clin Pharmacol. 1991;32 :31 –37
Hartley R, Green M, Quinn M, Levene MI. Pharmacokinetics of morphine infusion in premature neonates. Arch Dis Child. 1993;69 :55 –58
Bhat R, Chari G, Gulati A, Aldana O, Velamati R, Bhargava H. Pharmacokinetics of a single dose of morphine in preterm infants during the first week of life. J Pediatr. 1990;117 :477 –481
Scott CS, Riggs KW, Ling EW, et al. Morphine pharmacokinetics and pain assessment in premature newborns. J Pediatr. 1999;135 :423 –429
Zahorodny W, Rom C, Whitney W, et al. The neonatal withdrawal inventory: a simplified score of newborn withdrawal. J Dev Behav Pediatr. 1998;19 :89 –93
Cunningham N. Inclusion of the nonverbal patient: a matter of moral emergency. Pain Forum. 1999;8 :110 –111
International Neonatal Network. The CRIB (Clinical Risk Index for Babies) score: a tool for assessing initial neonatal risk and comparing performance of neonatal intensive care units. Lancet. 1993;342 :193 –198
Wood CM, Rushforth JA, Hartley R, Dean H, Wild J, Levene MI. Randomized double blind trial of morphine versus diamorphine for sedation of preterm infants. Arch Dis Child Fetal Neonatal Ed. 1998;79 :F34 –F39
Jacqz-Aigrain E, Daoud P, Burtin P, Desplanques L, Beaufils F. Placebo-controlled trial of midazolam sedation in mechanically ventilated newborn babies. Lancet. 1994;344 :646 –650
Bhandari V, Narang A. Role of morphine or triclofos in neonatal ventilation: a pilot study . Bull Natl Neonatol Forum. 1992;6 :8 –9
Bhandari A, Bhandari V. Pathogenesis, pathology, and pathophysiology of pulmonary sequelae of bronchopulmonary dysplasia in premature neonates. Front Biosci. 2003;8 :e370 –e380
Miller J, Law AB, Parker RA, Sundell H, Silberberg AR, Cotton RB. Effects of morphine and pancuronium on lung volume and oxygenation in premature infants with hyaline membrane disease. J Pediatr. 1994;125 :97 –103
Alexander SM, Todres ID. The use of sedation and muscle relaxation in the ventilated infant. Clin Perinatol. 1998;25 :63 –78
Ng E, Taddio A, Ohlsson A. Intravenous midazolam infusion for the sedation of infants in the neonatal intensive care unit. Cochrane Database Syst Rev. 2000;(2) : CD002052
Simons SH, van Dijk M, van Lingen RA, et al. Routine morphine infusion in preterm newborns who received ventilatory support: a randomized controlled trial. JAMA. 2003;290 :2419 –2427
Quinn MW, Vokes A. Effect of morphine on respiratory drive in trigger ventilated pre-term infants. Early Hum Dev. 2000;59 :27 –35
Corrupt PA, Testa A, Bechalany A, El Tayar N, Descas P, Perrisoud D. Metabolites of morphine as molecular chameleons. J Med Chem. 1991;34 :1272 –1275
Christrup LL. Morphine metabolites. Acta Anaesth Scand. 1997;41 :116 –122
Osborne RJ, Joel SP, Trew D, Slevin M. Analgesic activity of morphine-6-glucuronide. Lancet. 1988;1 :828
Inturrisi CE. Morphine-6-glucuronide: a potent μ-agonist. Life Sci. 1987;41 :2845 –2849
Suzuki N, Kalso E, Rosenberg PH. Intrathecal morphine-3-glucuronide does not antagonise spinal anti-nociception by morphine or morphine-6-glucuronide in rats. Eur J Pharmacol. 1993;249 :247 –250
Smith MT, Watt JA, Cramond T. Morphine-3-glucuronide: a potent antagonist of morphine analgesia. Life Sci. 1990;47 :579
Bouwmeester NJ, van den Anker JN, Hop WC, Anand KJ, Tibboel D. Age- and therapy-related effects on morphine requirements and plasma concentrations of morphine and its metabolites in postoperative infants. Br J Anaesth. 2003;90 :642 –652
Bouwmeester NJ, Hop WC, van Dijk M, Anand KJ, van den Anker JN, Tibboel D. Postoperative pain in the neonate: age-related differences in morphine requirements and metabolism. Intensive Care Med. 2003;29 :2009 –2015
Kaaresen PI, Dùhlen G, Fundingsrud HP, Dahl LB. The use of CRIB (Clinical Risk Index for Babies) score in auditing the performance of one neonatal intensive care unit. Acta Paediatr. 1998;87 :195 –200
Rautonen J, Mkel A, Boyd H, Apajasalo M, Pohjavuori M. CRIB and SNAP: assessing the risk of death for preterm infants. Lancet. 1994;343 :1272 –1273(Vineet Bhandari, MD, DM, )
Karolinska Institute, Neonatal Research Unit, Astrid Lindgren's Children's Hospital, Stockholm, Sweden
Maryland Medical Research Institute, Baltimore, Maryland
Department of Pediatrics, University of Arkansas for Medical Sciences, Arkansas Children's Hospital, Little Rock, Arkansas
ABSTRACT
Background. The use of opioid therapy for sedation and analgesia among ventilated infants varies among care providers. The impact of opioid therapy early in the neonatal course of respiratory distress syndrome (RDS) on pulmonary outcomes is not known.
Objective. We tested the hypothesis that preterm neonates randomized to the morphine infusion group would have improved ventilatory outcomes, measured as shorter durations of ventilator or oxygen therapy, fewer air leaks, and lower incidence of bronchopulmonary dysplasia.
Methods. All 898 subjects (gestational age [GA] of 23 to 32 weeks) who were enrolled in the Neurologic Outcomes and Preemptive Analgesia in Neonates (NEOPAIN) trial formed the study cohort (morphine: 449 patients; placebo: 449 patients). Subjects received the masked study drug until they were weaned from the ventilator or for 14 days, whichever occurred earlier. Outcome measures included air leaks, duration of ventilation or oxygen therapy, hospitalization, bronchopulmonary dysplasia, and death.
Results. Subjects in the 2 groups had similar baseline characteristics (mean ± SD, morphine versus placebo: GA: 27.3 ± 2.3 vs 27.4 ± 2.3 weeks; birth weight: 1037 ± 340 vs 1054 ± 354 g). Infants in the morphine group required ventilator therapy significantly longer, compared with the placebo group (median [interquartile range]: 7 days [4–20 days] vs 6 days [3–19 days]). This difference in ventilation duration was significant for infants with GA of 27 to 29 weeks (6 days [4–12 days] vs 5 days [2–9 days]) and 30 to 32 weeks (4 days [3–6 days] vs 3 days [2–5 days]). Infants who received additional analgesia with intermittent morphine doses in both groups were sicker than those who were not given open-label morphine. After adjustment for birth weight, Clinical Risk Index for Babies scores, maternal chorioamnionitis, RDS requiring surfactant, and patent ductus arteriosus in a logistic regression model, the use of additional analgesia with morphine was associated independently with increased air leaks and longer durations of high-frequency ventilation, nasal continuous positive airway pressure, and oxygen therapy.
Conclusions. Morphine infusions do not improve short-term pulmonary outcomes among ventilated preterm neonates. Additional morphine doses were associated with worsening respiratory outcomes among preterm neonates with RDS.
Key Words: pain bronchopulmonary dysplasia lung newborn infant morphine
Abbreviations: AA, additional analgesia BPD, bronchopulmonary dysplasia BW, birth weight CPAP, continuous positive airway pressure CRIB, Clinical Risk Index for Babies GA, gestational age HFV, high-frequency ventilation M3G, morphine-3-glucuronide M6G, morphine-6-glucuronide nCPAP, nasal continuous positive airway pressure NEOPAIN, Neurologic Outcomes and Preemptive Analgesia in Neonates NOPAIN, Neonatal Outcome and Prolonged Analgesia in Neonates PDA, patent ductus arteriosus RDS, respiratory distress syndrome
The relief of pain per se is certainly an appropriate clinical goal for premature neonates.1 There is, however, significant variation in the use of opioid analgesics in NICUs.2 The impact of opioid therapy, used early in the neonatal course, on short-term outcomes is not known.3 The studies4–7 that have evaluated the impact of morphine on neonatal ventilatory outcomes have varied in their findings. Dyke at al4 reported that morphine therapy was associated with improved ventilator synchrony and reduced heart rate, respiratory rate, and duration of oxygen therapy among ventilated preterm neonates. That study, however, had only 26 infants enrolled and the results of that small pilot study should be interpreted with caution.4 Our pilot study, the Neonatal Outcome and Prolonged Analgesia in Neonates (NOPAIN) trial, suggested improved neurologic outcomes for the group of patients who received morphine (vs midazolam or placebo) but no differences in the number of days required for mechanical ventilation, continuous positive airway pressure (CPAP), or oxygen therapy.5
Among 20 ventilated premature infants randomized to receive fentanyl or placebo infusions, those who received fentanyl required a longer duration of mechanical ventilation but showed no effects on bronchopulmonary dysplasia (BPD), intraventricular hemorrhage, or sepsis.6 In another study of 27 infants (compared with 28 untreated control subjects), there were no significant differences in the short-term ventilatory parameters between the 2 groups.7 However, these studies4–7 had small numbers of subjects and inadequate power to detect or to reject any difference in these outcomes. Therefore, there is no reliable evidence regarding the effects of morphine on pulmonary outcomes among neonates.
The subjects enrolled in the Neurologic Outcomes and Preemptive Analgesia in Neonates (NEOPAIN) multicenter trial provided the opportunity to explore the effects of opioid therapy on pulmonary outcomes. We proposed, a priori, to test the hypothesis that preterm infants with respiratory distress syndrome (RDS) who received morphine infusions would have shorter durations of ventilation, CPAP, or oxygen therapy and decreased rates of pulmonary air leaks or BPD.
METHODS
Study Group
Methodologic details of the NEOPAIN trial were published recently.8 Briefly, preterm neonates who were born between 23 and 32 weeks of gestation, intubated before 72 hours of age, and ventilated for <8 hours at the time of enrollment were eligible. Neonates with major congenital anomalies, birth asphyxia, intrauterine growth restriction (<5th percentile), or maternal opioid addiction and those participating in other clinical trials were excluded. Written parental consent was obtained for all study infants. The NEOPAIN protocol and consent form were approved by local ethics committees at each of the 16 participating NICUs and by an external ethics committee at the coordinating center. An independent data and safety monitoring board was constituted to monitor the trial.
Randomization was performed with an automated telephone response system and was stratified according to the participating NICU and gestational age (GA) (23–26 weeks, 27–29 weeks, or 30–32 weeks), to ensure equal numbers in the morphine and placebo groups. To eliminate clinical bias, all clinical personnel were blinded to the study drug code, and unblinding of the treatment code (n = 6) was limited by stringent criteria.
Therapeutic Management
Data Collection
Data collection included baseline clinical and demographic characteristics (Table 1), vital signs, and clinical outcomes related to ventilator management. The clinical factors defined below were selected a priori, on the basis of clinically reported associations with mechanical ventilation or BPD.
Definitions
Maternal Variables
A positive culture from the amniotic fluid was defined as chorioamnionitis. Chorioamnionitis was also diagnosed on the basis of positive Gram stain results, positive white blood cell counts (50 cells per mm3), or low glucose concentrations (14 mg/dL) in the amniotic fluid. Chorioamnionitis was suspected if any 2 of the following clinical symptoms were present: maternal fever, leukocytosis, uterine tenderness, foul-smelling amniotic fluid, or fetal tachycardia.
Neonatal Variables
RDS was diagnosed on the basis of the clinical findings of respiratory distress and chest radiographs showing a diffuse reticulogranular pattern with air bronchograms. The Clinical Risk Index for Babies (CRIB), which includes birth weight (BW), GA, congenital malformations, maximal base excess, and minimal and maximal inspired oxygen concentrations in the first 12 hours, was calculated.17 Durations of mechanical ventilation and/or nasal CPAP (nCPAP) were defined as the completed days of these therapies. The number of doses of surfactant given for the treatment of RDS served as a surrogate measure of the severity of RDS. Patent ductus arteriosus (PDA) was diagnosed on the basis of clinical signs of high-output congestive heart failure and was confirmed with echocardiographic evidence of ductal patency. Air leaks (pneumothorax, pneumomediastinum, pulmonary interstitial emphysema, pneumoperitoneum, pneumopericardium, or subcutaneous emphysema) were diagnosed on the basis of chest radiographic findings, ultrasonographic findings, or clinical signs. Early-onset sepsis was defined as a positive blood culture in the first 72 hours with clinical signs of sepsis or strong clinical evidence of sepsis with abnormal hematologic values but no positive culture. BPD was diagnosed on the basis of clinical and chest radiographic findings of BPD and the need for supplemental oxygen at a corrected GA of 36 weeks. Length of stay was defined as the number of completed days in the hospital.
Statistical Analyses
The hypotheses, study protocol, sample size, and methods for this prospective study were determined a priori before the initiation of statistical analyses. Assuming an incidence of BPD in the target population (GA of 23–32 weeks) of 30%, we estimated that 370 infants in each group would allow us to detect a change of at least 30% at = .05 and a power of 0.8. Treatment group demographic features and outcomes were compared with Student's t test, 2 tests, or, for contingency tables with small cell numbers, Fisher's exact tests. For data that did not have a strictly normal distribution, the results are presented as median and interquartile range, and nonparametric tests (Wilcoxon test and Kruskal-Wallis test) were used for analyses (Tables 1–4), as appropriate.
Adjusting for BW and CRIB scores, we analyzed the effects of AA (intermittent morphine boluses) and clinical factors (maternal chorioamnionitis, RDS requiring surfactant, and PDA) on the days of ventilation, nCPAP, and oxygen treatment, as well as the length of stay, with multiple linear regression models (Table 5). Each predictor factor was entered into the models as a binary (0 or 1) variable, with 1 indicating the presence of the factor and 0 indicating the absence of the factor. The CRIB score, with a range of 0 to 20, was entered into the models as a ranking variable, with 10 as the reference score. Results of the multiple linear regression models are presented as parameter estimates and SEs.
Factors associated with the outcomes of pulmonary air leaks, BPD, and BPD or death were analyzed with logistic models (Table 5). The fit of each logistic model was assessed with the Hosmer-Lemeshow goodness-of-fit test, and the global test that all regression parameters were 0 was tested with the –2 log-likelihood statistic. Results of these analyses are presented as odds ratios with 2-sided 95% confidence intervals, to show the effect of each predictor variable on the indicated outcome. Results are presented for covariates that were statistically significant. All analyses were performed with SAS statistical software (SAS Institute, Cary, NC), with the critical P value set at .05.
RESULTS
Data on 898 infants were available for analysis, with 449 in the morphine group and 449 in the placebo group. Demographic data are shown in Table 1. The 2 groups were well matched for maternal and infant factors (Table 1).
The neonatal outcomes of all infants are presented in Table 2. The frequency of RDS and the overall severity of illness (CRIB scores) were similar in the 2 groups. Significantly more days were spent on mechanical ventilation in the morphine group, compared with the placebo group (P < .01), as reported earlier.8 No differences in the durations of nCPAP, oxygen therapy, or hospitalization or in the incidences of air leaks, early-onset sepsis, total sepsis, postnatal steroid use, BPD, or deaths occurred between the 2 randomized groups (Table 2).
In the 3 different GA strata (23–26 weeks, 27–29 weeks, and 30–32 weeks), there were no differences in the incidence of RDS, number of surfactant doses, incidence of PDA, indomethacin use, incidence of early-onset or total sepsis, length of stay, or number of deaths between the morphine and placebo groups (data not shown). In the 23 to 26 weeks' GA category, there were no differences between the 2 groups in the total ventilation days (median [interquartile range]: 18 days [7–39 days] vs 20 days [6–37.5 days]; P = .94) or duration of oxygen use (55 days [11–83 days] vs 55 days [18–78 days]; P = .69). The incidences of air leaks (21% vs 19%) and BPD (38% vs 36%) were not significantly different in the morphine and placebo arms.
In the 27 to 29 weeks' GA category, the morphine group required longer ventilation than did the placebo group (6 days [4–12 days] vs 5 days [2–9 days]; P < .01), but no difference was noted in the duration of oxygen use (31 days [13–52 days] vs 34 days [14–49.5 days]; P = .95). The incidences of air leaks (7% vs 6%) and BPD (17% vs 18%) were not significantly different in the morphine and placebo arms. In the 30 to 32 weeks' GA category, the morphine group required longer ventilation than did the placebo group (4 days [3–6 days] vs 3 days [2–5 days]; P = .02), but no difference was noted in the duration of oxygen use (8 days [4–25 days] vs 7.5 days [3–15 days]; P = .25). The incidences of air leaks (5% vs 10%) and BPD (6% vs 7%) were not significantly different in the morphine and placebo arms. Differences in the duration of ventilation remained significant for the 27 to 29 weeks' GA (6 days [4–12 days] vs 5 days [2–9 days]; P < .01) and 30 to 32 weeks' GA (4 days [3–6 days] vs 3 days [2–5 days]; P = .03) strata, even after exclusion of deaths from both randomized groups.
To evaluate the effect of open-label AA in the morphine and placebo groups, data were analyzed in each group by comparing infants who did versus did not receive AA. Their demographic data are noted in Table 3. Infants who received AA, in both randomized groups, had significantly lower maternal age, less maternal chorioamnionitis, lower GA and BW, and lower Apgar scores at 1 and 5 minutes (Table 3).
A higher incidence of RDS and greater severity of illness (CRIB scores) were noted for infants who received AA (Table 4). Therefore, infants who received AA in either randomized group were "sicker" and had greater physiologic instability. This might have led to increased complications with more frequent procedures, resulting in increased assessments of pain by their caregivers and consequently increased use of AA. Table 4 shows significantly higher incidences of PDA, air leaks, and chest tubes and worse respiratory outcomes, with longer durations of ventilation, oxygen therapy, and hospital stay, higher incidence of BPD, and more neonatal deaths, among infants who received AA, compared with those who did not.
To examine whether morphine therapy contributed to the respiratory outcomes outlined above, we adjusted for BW, CRIB scores (which include GA in the composite scores), maternal chorioamnionitis, RDS requiring surfactant, and PDA in our logistic regression models, using placebo/no AA as the reference group. Preemptive morphine infusions did not alter these outcomes, but the use of morphine AA continued to be a strong independent predictor of worse respiratory outcomes even after adjustment for the aforementioned factors (Table 5).
Infants who received AA in the placebo group had a 3.4-fold increase in pulmonary air leaks (P < .01), compared with infants who received placebo and no AA. On average, these infants spent an additional 2.6 days on HFV (P < .01) and 3.2 days on nCPAP (P = .02) and required supplemental oxygen for an additional 7.2 days (P = .01), compared with infants who received placebo and no AA (Table 5).
Infants who received AA in the morphine group had even worse respiratory outcomes (Table 5), with a 4.3-fold increase in air leaks (P < .01), compared with infants who received placebo and no AA. On average, infants who received AA in the morphine group spent an additional 6.7 days on positive-pressure ventilation (P < .01), with similar contributions from conventional ventilation (P = .04) and HFV (P < .01). These infants also spent more time on nCPAP (additional 2.9 days; P = .04) and on supplemental oxygen (additional 8.1 days; P < .01).
DISCUSSION
A previous study suggested that morphine infusions increase the synchronicity of spontaneous and ventilator-derived breaths among preterm infants with RDS and reduce the duration of oxygen therapy.4 In such a scenario, if morphine analgesia decreases the ventilatory pressures and need for oxygen, then potentially it could reduce ventilator-induced lung injury, perhaps resulting in a decreased incidence of BPD.
Results from the NEOPAIN multicenter trial suggest that use of preemptive morphine analgesia among premature infants with RDS does not decrease the incidences of pulmonary air leaks, BPD, or other respiratory outcomes. In a randomized, double-blind trial of morphine versus diamorphine among ventilated preterm neonates (n = 88), the 2 drug regimens were equally effective in terms of sedation, with no significant differences in mortality rates, ventilator days, or BPD rates.18 In the NOPAIN pilot study, no significant differences in ventilatory outcomes among the morphine (n = 24), midazolam (n = 22), and placebo (n = 21) groups were noted, but the durations of ventilation, nCPAP, and oxygen therapy were lowest in the morphine group.5
Ventilatory indices were unchanged in a short-term study comparing ventilated infants with RDS who received midazolam (n = 24) versus placebo (n = 22).19 Among mechanically ventilated neonates with RDS who did not receive surfactant, infants who received morphine (n = 11) had improved outcomes, compared with those who received chloral hydrate (n = 8), as evidenced by lower mean airway pressures, ventilator rates, and need for oxygen and improved survival rates in the morphine group.20
In our study, overall there was no difference in the incidence of BPD, despite significant differences in the duration of ventilation (Table 2). This highlights the complex and multifactorial nature of BPD.21 It is noteworthy that the incidence of BPD was higher among the infants who received more morphine (Table 4). Interestingly, after correction for variables known to affect BPD, AA with morphine continued to be associated significantly with longer duration of ventilation and oxygen therapy but not with the incidence of BPD (Table 5). Therefore, traditional environmental factors (for example, ventilator-induced lung injury, hyperoxia, sepsis, and PDA) may not be solely responsible and may require association with other (for example, genetic) influences to cause the disordered lung development that is the hallmark of BPD.21
On the basis of currently available physiologic/behavioral markers, opioid analgesia seems definitely to reduce pain/stress among ventilated preterm neonates.5–7,18 In terms of pulmonary outcomes, previous data suggested that opioid analgesia may be preferable over sedation,5,20 but the combination of opioid analgesia and muscle relaxation and/or sedative agents should be used with caution among ventilated neonates.22,23 Furthermore, a meta-analysis has raised concerns regarding the use of midazolam infusions among neonates.24
Most studies examining respiratory outcomes among ventilated neonates receiving analgesia enrolled a small number of infants5–7,19,20 and/or lacked a placebo arm.3,18,20 A recent randomized trial of morphine (n = 77) versus placebo (n = 73) among premature ventilated infants did not show any differences in the duration of ventilation or the incidence of BPD between the 2 groups.25 The large cohort of infants in the NEOPAIN trial (with a placebo arm) provided a more definitive opportunity to assess the effects of opioid analgesia on the respiratory outcomes of ventilated preterm infants.
Infants receiving morphine infusions were ventilated for significantly longer periods than were those receiving placebo, with differences primarily among 27- to 29-week neonates (receiving 20 μg/kg per hour morphine) and 30- to 32-week neonates (receiving 30 μg/kg per hour morphine). This most likely reflects the respiratory depression caused by increased doses of morphine, compared with that for 23- to 26-week neonates, who received 10 μg/kg per hour morphine. It is possible that the doses of morphine administered in the present study were excessive for the population studied. This could be a result of the dosing regimen or drug metabolism (see below). Weaning of the study drug was performed only for factors such as extreme suppression of the respiratory drive, and it is quite likely that the effects of the morphine infusion included not only analgesia but also excessive sedation. The results indicating a longer duration of mechanical ventilation in the morphine group would then be logical, because infants in the morphine group would be more sedated and therefore would have a lower respiratory drive.
The use of infant-triggered respiratory rate as a measure of respiratory drive is a novel approach to studying the problem of respiratory depression with morphine, although it is dependent on the infant's interaction with the ventilator and the type of triggering devices in the ventilator.26 Among trigger-ventilated preterm neonates (n = 14; BW: 1.37 kg; GA: 30 weeks) who received morphine (100 μg/kg bolus and 10 μg/kg/hour infusion), infants who had morphine-6-glucuronide (M6G) detectable at 12 hours (n = 6) showed a significantly greater reduction in triggered breath rate (–22 beats per minute), compared with those who did not (n = 3; –4 beats per minute) (P = .03).26 This raises questions about the role of metabolites of morphine in producing its clinical effects. Morphine is metabolized in the liver to the polar conjugates morphine-3-glucuronide (M3G) and M6G. The formation of the glucuronides is slow in preterm newborns because of immature liver function, but their half-lives are prolonged because of poor renal excretion. Despite their poor lipophilicity, there is clear evidence of brain penetration by the glucuronides,27,28 and this is likely to occur among preterm newborns with a poorly developed blood-brain barrier. There is evidence that M6G is a more potent analgesic agent and respiratory depressant than morphine itself,29,30 whereas M3G has antiopioid effects, although this evidence is not consistent.31,32 The formation of M3G in preterm newborns occurs earlier and in greater amounts than that of M6G.26 Therefore, in assessments of the respiratory depressant effects of morphine among preterm neonates, variables such as hepatic metabolism and renal excretion may be more important than the dosing regimen.
In a recent double-blind study comparing continuous infusion of 10 μg/kg per hour versus intermittent morphine boluses of 30 μg/kg every 3 hours among term neonates (and older children), it was reported that neonates had a narrower therapeutic window for postoperative morphine analgesia than did older age groups, with no difference in the safety or effectiveness of intermittent doses, compared with continuous infusions.33 The high plasma concentrations of morphine metabolites in the neonates were the result of low renal clearance, which was confirmed by the significant correlation between serum creatinine and M6G levels in this age group. The decreased clearance of morphine and M6G explained the increased analgesic effects among neonates.33 The authors concluded that morphine given intermittently does not provide any clinical advantages and that a continuous morphine infusion is probably safer for neonates.33 Another report from the same investigators described the effects of age and other clinical factors among 68 neonates (52 who were <7 days of age and 16 who were 7 days of age) after major surgery.34 Although pain scores were no different, the younger neonates had different morphine requirements (10 vs 10.8 μg/kg per hour), plasma morphine concentrations (23.0 vs 15.3 ng/mL), and M6G/morphine ratios (0.6 vs 1.5).34 Respiratory insufficiency occurred for 5 neonates who received intermittent morphine, but this difference between the groups was not significant. The authors concluded that neonates 7 days of age required significantly less morphine postoperatively than older neonates.34
These data suggest that patients of younger GA and sicker infants (requiring mechanical ventilation) would have lower requirements for morphine and that intermittent boluses of morphine might have more adverse effects. It is noteworthy that the infants of youngest GA (23–26 weeks) in our study were receiving 10 μg/kg per hour, the same dose as the term neonates in the aforementioned 2 reports.33,34
Another important result, apart from prolonged ventilation associated with a higher dose of morphine infusion, was the fact that morphine AA was an independent predictor of worse neonatal respiratory outcomes, even after controlling for the effects of BW and severity of illness (CRIB score). The CRIB score was superior to BW35 or the Score for Neonatal Acute Physiology for predicting survival to discharge.36
Interestingly, logistic regression analyses showed that the respiratory outcomes of infants who received preemptive morphine infusions alone (no AA) were not significantly different from those of the placebo/no AA group. In contrast, infants who received intermittent doses of morphine (AA) had worse respiratory outcomes, regardless of whether they received placebo or morphine infusions (Table 5). This effect was somewhat accentuated in the morphine group with AA, which suggests a cumulative dose effect of morphine on neonatal respiratory outcomes. This association of AA with worse pulmonary outcomes does not imply causation, because other unknown and/or unrecorded factors might be contributing to the results.
Because morphine AA was given at the discretion of the caregiver (based on their evaluation of "pain/discomfort" in the neonate), our data suggest that, for extremely sick, ventilated, premature newborns, providing "pain relief" through intermittent boluses of morphine might worsen their outcomes. We cannot discount completely the contribution of disease severity in this population to worse respiratory outcomes, but intermittent boluses of morphine seem to worsen neonatal ventilatory outcomes independently. We suggest caution in the use of intermittent morphine boluses for treating "pain/discomfort" in such populations.
CONCLUSIONS
Our data suggest that use of morphine analgesia for ventilated preterm infants does not reduce the incidence of BPD or other respiratory outcomes. Morphine infusions of >10 μg/kg per hour might prolong the period of ventilation for preterm neonates, compared with control subjects. Furthermore, intermittent boluses of morphine were associated with worsening respiratory outcomes among ventilated preterm neonates.
ACKNOWLEDGMENTS
This study was supported by grants from the National Institute for Child Health and Human Development (HD36484 to K.J.S.A. and HD36270 to B.A.B.), from the Chief Scientist's Office of the Scottish Executive (to N. McIntosh), from the Vardal Foundation, Free Masons, and Swedish Medical Association, Sweden (to H. Lagercrantz and L.L.B.), from the Fondation pour la Santé CNP, France (to R. Carbajal and R. Lenclen), and from the rebro University Hospital Research Foundation, Sweden (to J. Schollin and M. Eriksson).
Centers and participants in the NEOPAIN trial were: United States: Arkansas Children's Hospital (Little Rock, AR), K.J.S. Anand, J. Seibert, M.B. Moore, H. Farrar, J. Valentine, T.J. Green, L. Letzig; Maryland Medical Research Institute (Baltimore, MD), B.A. Barton, S.S. Kronsberg, S. Fick; University of Arkansas for Medical Sciences (Little Rock, AR), R.W. Hall, R. Arrington, R. Smith; University of Kentucky (Lexington, KY), N. Desai, V.L. Whitehead, J. L. Sampers, P.E. DeFranco, L.A. Shook, T.H. Pauly; Tufts University and New England Floating Hospital, B.A. Shephard, B. Mackinnon, J. Fiascone; Wake Medical Center (Raleigh, NC), T.E. Young, K. Carr, M.R. Johnson, K. McDaniel; Mercy Hospital and Medical Center (Chicago, IL), R. Vasa, J.S. Teji, D. Jahn; University of Pennsylvania and Pennsylvania Hospital (Philadelphia, PA), V.K. Bhutani, J. R. Gerdes, S. Abbasi, M.K.S. Grous, A. Scwhoebel; Medical Center of Delaware, D.J. Tuttle, K.H. Leef, J.L. Stefano; University of Illinois at Chicago (Chicago, IL), R. Bhat, T.N.K. Raju, G. Chari; Albert Einstein Medical Center (Philadelphia, PA), V. Bhandari, H. Hurt, J. Giannetta; Brigham and Women's Hospital (Boston, MA), E.C. Eichenwald, K. Housefield; Long Beach Memorial Hospital (Long Beach, CA), G. Padilla, K. Norris; France: Centre Hospitalier Poissy Saint Germain, Site Poissy, R. Carbajal, R. Lenclen, M. Jugie; Sweden: Karolinska Institute at Astrid Lindgren's Children's Hospital (Stockholm, Sweden), H. Lagercrantz, L.L. Bergqvist; rebro University Hospital (rebro, Sweden), J. Schollin, M. Eriksson; United Kingdom: Simpson Memorial Maternity Pavilion (Edinburgh, Scotland), N. McIntosh, G. Menon, E.M. Boyle; Data and Safety Monitoring Board: R. Troug (Harvard Medical School and Children's Hospital, Boston, MA), A. Aynsley-Green (Institute of Child Health and Hospital for Sick Children, London, United Kingdom), K.D. Craig (University of British Columbia, Vancouver, Canada), C.C. Johnston (McGill University School of Nursing and Montreal Children's Hospital, Montreal, Canada), M. Walden (Baylor University and Texas Children's Hospital, Houston, TX); W.A. Silverman (Greenbrae, CA); E. Bancalari (University of Miami, FL); J. Rowe (National Institutes of Health, Bethesda, MD, and Office on Women's Health, Washington DC).
We thank all of the physicians, nurses, pharmacists, ultrasonographers, occupational therapists, and physical therapists at the participating institutions and the parents who gave consent for this study.
FOOTNOTES
Accepted Nov 24, 2004.
No conflict of interest declared.
REFERENCES
Anand KJS. Clinical importance of pain and stress in preterm neonates. Biol Neonate. 1998;73 :1 –9
Kahn DJ, Richardson DK, Gray JE, et al. Variation among neonatal intensive care units in narcotic administration. Arch Pediatr Adolesc Med. 1998;152 :844 –851
Kennedy KA, Tyson JE. Narcotic analgesia for ventilated newborns: are placebo-controlled trials ethical and necessary J Pediatr. 1999;134 :127 –129
Dyke MP, Kohan R, Evans S. Morphine increases synchronous ventilation in preterm infants. J Paediatr Child Health. 1995;31 :176 –179
Anand KJS, McIntosh N, Lagercrantz H, et al. Analgesia and sedation in preterm neonates who require ventilatory support: results from the NOPAIN trial. Arch Pediatr Adolesc Med. 1999;153 :331 –338
Orsini AJ, Leef KH, Costarino A, Dettore MD, Stefano JL. Routine use of fentanyl infusions for pain and stress reduction in infants with respiratory distress syndrome. J Pediatr. 1996;129 :140 –145
Lago P, Benini F, Agosto C, Zacchello F. Randomised controlled trial of low dose fentanyl infusion in preterm infants with hyaline membrane disease. Arch Dis Child Fetal Neonatal Ed. 1998;79 :F194 –F197
Anand KJS, Hall RW, Desai NS, et al. Effect of morphine analgesia in ventilated preterm neonates: primary outcomes from the NEOPAIN randomized trial. Lancet. 2004;263 :1673 –1682
Bhat R, Abu-Harb M, Chari G, Gulati A. Morphine metabolism in acutely ill preterm newborn infants. J Pediatr. 1992;120 :795 –799
Barrett DA, Barker DP, Rutter N, Pawula M, Shaw PN. Morphine, morphine-6-glucuronide and morphine-3-glucuronide pharmacokinetics in newborn infants receiving diamorphine infusions. Br J Clin Pharmacol. 1996;41 :531 –537
Barrett DA, Elias-Jones AC, Rutter N, Shaw PN, Davis SS. Morphine kinetics after diamorphine infusion in premature neonates. Br J Clin Pharmacol. 1991;32 :31 –37
Hartley R, Green M, Quinn M, Levene MI. Pharmacokinetics of morphine infusion in premature neonates. Arch Dis Child. 1993;69 :55 –58
Bhat R, Chari G, Gulati A, Aldana O, Velamati R, Bhargava H. Pharmacokinetics of a single dose of morphine in preterm infants during the first week of life. J Pediatr. 1990;117 :477 –481
Scott CS, Riggs KW, Ling EW, et al. Morphine pharmacokinetics and pain assessment in premature newborns. J Pediatr. 1999;135 :423 –429
Zahorodny W, Rom C, Whitney W, et al. The neonatal withdrawal inventory: a simplified score of newborn withdrawal. J Dev Behav Pediatr. 1998;19 :89 –93
Cunningham N. Inclusion of the nonverbal patient: a matter of moral emergency. Pain Forum. 1999;8 :110 –111
International Neonatal Network. The CRIB (Clinical Risk Index for Babies) score: a tool for assessing initial neonatal risk and comparing performance of neonatal intensive care units. Lancet. 1993;342 :193 –198
Wood CM, Rushforth JA, Hartley R, Dean H, Wild J, Levene MI. Randomized double blind trial of morphine versus diamorphine for sedation of preterm infants. Arch Dis Child Fetal Neonatal Ed. 1998;79 :F34 –F39
Jacqz-Aigrain E, Daoud P, Burtin P, Desplanques L, Beaufils F. Placebo-controlled trial of midazolam sedation in mechanically ventilated newborn babies. Lancet. 1994;344 :646 –650
Bhandari V, Narang A. Role of morphine or triclofos in neonatal ventilation: a pilot study . Bull Natl Neonatol Forum. 1992;6 :8 –9
Bhandari A, Bhandari V. Pathogenesis, pathology, and pathophysiology of pulmonary sequelae of bronchopulmonary dysplasia in premature neonates. Front Biosci. 2003;8 :e370 –e380
Miller J, Law AB, Parker RA, Sundell H, Silberberg AR, Cotton RB. Effects of morphine and pancuronium on lung volume and oxygenation in premature infants with hyaline membrane disease. J Pediatr. 1994;125 :97 –103
Alexander SM, Todres ID. The use of sedation and muscle relaxation in the ventilated infant. Clin Perinatol. 1998;25 :63 –78
Ng E, Taddio A, Ohlsson A. Intravenous midazolam infusion for the sedation of infants in the neonatal intensive care unit. Cochrane Database Syst Rev. 2000;(2) : CD002052
Simons SH, van Dijk M, van Lingen RA, et al. Routine morphine infusion in preterm newborns who received ventilatory support: a randomized controlled trial. JAMA. 2003;290 :2419 –2427
Quinn MW, Vokes A. Effect of morphine on respiratory drive in trigger ventilated pre-term infants. Early Hum Dev. 2000;59 :27 –35
Corrupt PA, Testa A, Bechalany A, El Tayar N, Descas P, Perrisoud D. Metabolites of morphine as molecular chameleons. J Med Chem. 1991;34 :1272 –1275
Christrup LL. Morphine metabolites. Acta Anaesth Scand. 1997;41 :116 –122
Osborne RJ, Joel SP, Trew D, Slevin M. Analgesic activity of morphine-6-glucuronide. Lancet. 1988;1 :828
Inturrisi CE. Morphine-6-glucuronide: a potent μ-agonist. Life Sci. 1987;41 :2845 –2849
Suzuki N, Kalso E, Rosenberg PH. Intrathecal morphine-3-glucuronide does not antagonise spinal anti-nociception by morphine or morphine-6-glucuronide in rats. Eur J Pharmacol. 1993;249 :247 –250
Smith MT, Watt JA, Cramond T. Morphine-3-glucuronide: a potent antagonist of morphine analgesia. Life Sci. 1990;47 :579
Bouwmeester NJ, van den Anker JN, Hop WC, Anand KJ, Tibboel D. Age- and therapy-related effects on morphine requirements and plasma concentrations of morphine and its metabolites in postoperative infants. Br J Anaesth. 2003;90 :642 –652
Bouwmeester NJ, Hop WC, van Dijk M, Anand KJ, van den Anker JN, Tibboel D. Postoperative pain in the neonate: age-related differences in morphine requirements and metabolism. Intensive Care Med. 2003;29 :2009 –2015
Kaaresen PI, Dùhlen G, Fundingsrud HP, Dahl LB. The use of CRIB (Clinical Risk Index for Babies) score in auditing the performance of one neonatal intensive care unit. Acta Paediatr. 1998;87 :195 –200
Rautonen J, Mkel A, Boyd H, Apajasalo M, Pohjavuori M. CRIB and SNAP: assessing the risk of death for preterm infants. Lancet. 1994;343 :1272 –1273(Vineet Bhandari, MD, DM, )