Materials and Methods
A database that linked maternal and neonatal hospital discharge records to birth certificate vital statistics records was used. The linkage of vital statistics was established for all civilian hospitals that reported to the California Office of Statewide Health Planning and Development from January 1, 1994, to December 31, 1995. That database did not include births at military facilities, home deliveries, out-of-state deliveries, or births at birthing centers not reporting to the California office. The method was able to link 98.9% of maternal and 98.6% of neonatal hospital discharge records with the birth record vital statistics, an overall linkage of 97.9%. This generated a database of 1,094,298 deliveries. With the use of statistical software, the database was searched using International Classification of Diseases, 9th revision15 (ICD-9), codes, which resulted in specific data for statistical analysis.
The database was sampled for the diagnosis of brachial plexus injury and those records were examined for ICD-9 codes relating to pregnancy outcomes and maternal complications. The linked database was searched for demographic, antepartum, intrapartum, and post-partum diagnoses, including diabetes and operative vaginal delivery. Modes of delivery and stratified birth weights were compared with the frequency of brachial plexus injury. Unadjusted odds ratios (ORs) were calculated comparing for presence of brachial plexus injury. The frequency of the following birth outcomes were determined: birth trauma (ICD-9 767), birth asphyxia (ICD-9 768), fetal growth restriction (FGR, ICD-9 764), intraventricular hemorrhage (ICD-9 772.1), sub-arachnoid hemorrhage (ICD-9 772.2), and infant length of stay. Incidences of the following complications were determined from the maternal records: other malpresentation (ICD-9 763.1), shoulder dystocia (ICD-9 660.4), gestational diabetes (ICD-9 648.8), and prematurity. The terms were from ICD-915 and were not defined further.
Among 1,094,298 women who delivered during the 2 years, 1611 (0.15%) had diagnoses of brachial plexus injury recorded. The individual percentage risks of brachial plexus injury and associated ORs, separated by mode of delivery, are shown in Table 1. The risk of brachial plexus injury is greater in deliveries assisted by forceps or vacuum extractor. There were 60 cases (0.03%) of brachial plexus injury among women who delivered by cesarean; the frequency of brachial plexus injury among women who delivered vaginally was greater (Table 1). The risk of brachial plexus injury increased with increasing birth weight and particular modes of delivery (Figure 1). Forceps, vacuum extractor, and cesarean deliveries were associated with similar increases in brachial plexus injury with increasing birth weight.
The effect of gestational diabetes (OR 1.9, 95% confidence interval [CI] 1.7, 2.1) on brachial plexus injury is shown in Figure 2. Assisted vaginal deliveries (forceps and vacuum extractor) in nondiabetic women was equivalent to spontaneous vaginal deliveries in diabetic women in terms of frequency of brachial plexus injury (Figure 2). The macrosomic (> 4500 g) newborns of diabetic women who had assisted vaginal deliveries had the highest brachial plexus injury rate (7.8%). The frequency of shoulder dystocia and brachial plexus injury increased with increasing birth weight, with 22% of newborns weighing 2.5–3.5 kg with brachial plexus injury also having shoulder dystocia. This increased to 74% in newborns weighing more than 4.5 kg (Figure 3). Overall, 53% of brachial plexus injury cases involved diagnoses of shoulder dystocia. However, the frequency of diagnosis of other malpresentation (ICD-9 763.1) was increased in all birth weight categories (Figure 3), exceeding shoulder dystocia in the low birth weight category. Other maternal and neonatal outcomes are displayed in Table 2. All types of birth asphyxia were increased with brachial plexus injury. Prematurity and FGR appeared to be protective against brachial plexus injury.
We found individual and collective risk factors associated with diagnoses of brachial plexus injury in our population of more than 1 million deliveries. This large population allowed for a detailed analysis and reflects a true population frequency (0.15%) for the state of California.9–14 Brachial plexus injury increased with increasing birth weight, shoulder dystocia, assisted vaginal deliveries, and gestational diabetes. The highest risk group for having newborns with brachial plexus injury (7.8%) was diabetic women who underwent forceps or vacuum-assisted deliveries and whose infants weighed more than 4.5 kg at birth (Figure 2). The frequency of diagnosis of other malpresentation was increased (OR 73.6, 95% CI 66, 83) in our population, suggesting non-shoulder dystocia causes of brachial plexus injury. Vacuum extractor and forceps deliveries increased the risk of brachial plexus injury. Cesarean delivery decreased the risk of but did not prevent brachial plexus injury (Figure 1). Neonatal morbidity was increased in newborns in our population (Table 2). The majority of newborns were discharged home, a small percentage were transferred to other institutions or referred for home health services.
One finding in our study was that more than half (53%) of all cases of brachial plexus injury were associated with diagnoses of shoulder dystocia. Among macrosomic infants, shoulder dystocia was absent in 26% of cases of brachial plexus injury. Graham et al3 found that 53% of cases of brachial plexus injury also involved diagnoses of shoulder dystocia, whereas Jennett et al2 found even fewer (43%) cases of brachial plexus injury to be associated with shoulder dystocia. Not all cases of brachial plexus injury are the result of difficult deliveries or traction on the anterior shoulder.4,16,17 Ouzounian et al4 reported on eight infants without shoulder dystocia, four of whom had a brachial plexus injury in the posterior arm that did not undergo downward traction during delivery. Shoulder dystocia might be important in brachial plexus injury, especially in macrosomic infants, but it cannot explain completely brachial plexus injury in low- and normal-weight infants.
We were surprised to find an increase in the frequency of diagnosis of other malpresentation (ICD-9, 763.1), associated (OR 73.6, CI 66, 83) with brachial plexus injury less frequently only than shoulder dystocia (Table 2). The increase in frequency of diagnosis of other malpresentation was equivalent across birth weight categories, compared with shoulder dystocia, whose frequency of diagnosis increased with increasing birth weight (Figure 3). The other malpresentation diagnosis was coded when a fetus presented abnormally (other than breech) in labor or delivery, and this diagnosis was coded in 0.5% of the entire population. Exact fetal presentations cannot be determined from our data set because the ICD-9 coding for other malpresentation is recorded in the discharge summaries. Whether the increase in that diagnosis was due to a reporting or coding error cannot be determined from our data set. If the increase in frequency of other malpresentation is accurate, it might show a cause of brachial plexus injury. Our findings suggest two possibilities: 1) shoulder dystocia, resulting in excessive traction on the brachial plexus, is the cause of a certain percentage of cases of brachial plexus injury; and, 2) fetal malpresentation, before or during labor and delivery, results in brachial plexus injury. A small percentage (15%, Figure 3) of macrosomic infants with brachial plexus injury were not associated with shoulder dystocia or other malpresentation, suggesting other possible causes of brachial plexus injury.
A difficult problem when one is confronted with a newborn with brachial plexus injury, is whether the injury was preventable. If identifiable risk factors can be found, an elective cesarean delivery might prevent brachial plexus injury. The majority (92%) of patients in the highest risk group (diabetic women who underwent assisted vaginal delivery and whose infants weighed more than 4.5 kg at birth) did not have brachial plexus injury and cesarean delivery would have been unnecessary. If inaccuracy in estimating fetal weight is added to this equation, our success rate for identifying the high-risk group would be worse. Many associations were found between obstetric conditions (macrosomia, diabetes, forceps and vacuum extractor deliveries, obesity, prolonged labor) and brachial plexus injury, but predicting which newborns will have brachial plexus injuries is unreliable.6,7 Infants who are delivered by cesarean have decreased risks of brachial plexus injury.10,18 However, routine use of cesarean delivery in certain patients at increased risk was not warranted. Macrosomia is commonly associated with brachial plexus injury; however, Ecker et al6 could not recommend routine use of cesarean delivery in the case of macrosomia. Rouse et al19 found no benefit to elective cesarean delivery in women with estimated fetal weights of more than 4.5 kg, unless the patients also had diabetes in which case the policy (of elective cesarean) became tenable. Perlow et al20 used multiple logistic regression to determine whether brachial plexus injury could be predicted. Only 19% of cases of brachial plexus injury were able to be identified before delivery and thus possibly were prevented.20
The majority of cases of brachial plexus injury diagnosed at birth resolve shortly thereafter. Nocon et al7 found that 96% of 28 cases of brachial plexus injury diagnosed at birth resolved within 6 months after delivery. Morrison et al,8 in a 10-year review, reported that 91% of 82 cases of brachial plexus injury resolved with no sequelae. Others13,14,21 report similar resolution rates. The best correlation with permanent brachial plexus injury is transient brachial plexus injury, which, as we discussed earlier, cannot be identified reliably. Permanent brachial plexus injury itself cannot be identified reliably before birth. An important question about patient management involves women who present with histories of brachial plexus injury in previous pregnancies. Literature reporting on recurrent brachial plexus injury is limited. al-Qattan and al-Kharfy22 hand surgeons who treat patients with permanent brachial plexus injury, reported a high recurrence of brachial plexus injury in their patients. Of eight women whose children had prior permanent brachial plexus injury, who delivered a total of 16 more offspring, two-thirds had infants with permanent brachial plexus injury, and the injury was worse in newborns who were also macrosomic. al-Qattan and al-Kharfy22 concluded that elective cesarean delivery might be indicated in cases of macrosomia and a history of brachial plexus injury. Their population included a large percentage of patients with permanent brachial plexus injury, and whether the conclusion applies to patients with resolved brachial plexus injury (the majority of patients) is unclear. Gordon et al13 found that 14% of their 59 subjects with brachial plexus injury had histories of brachial plexus injury in previous pregnancies. That information, if acted on through performance of elective cesarean delivery could have prevented brachial plexus injury in those newborns. Offering elective cesarean delivery to women with children with permanent brachial plexus injury is indicated. In cases in which prior brachial plexus injury resolved after birth, cesarean delivery should be discussed and considered as an option. We were unable to determine whether the brachial plexus injuries reported were permanent or temporary. Only information on discharges from the hospital was available, not information on follow-up visits.
Neonatal morbidities were increased in the brachial plexus injury population, with an increase in mild and severe birth asphyxia and subarachnoid hemorrhage (Table 2), confirming previous publications.10,13,23 Evidence of increased morbidity was demonstrated indirectly by a three-fold increase in newborn lengths of stay, compared with newborns without brachial plexus injury (Table 2). Additional evidence of increased morbidity was shown by the increase in referrals to other hospitals and referrals for home health services. There were no reported neonatal deaths in the 1583 infants who were discharged from delivering hospitals. The outcomes of the 28 neonates who were referred to other hospitals are unknown. Prematurity and FGR were protective against brachial plexus injury, probably relating to lower birth weight in those infants. The data presented here show that there are potentially multiple causes of brachial plexus injury, including shoulder dystocia, malpresentation, diabetes, and operative vaginal delivery.
1. Jakobovits A. Medico-legal aspects of brachial plexus injury: The obstetrician's point of view. Med Law 1996;15:175–82.
2. Jennett RJ, Tarby TJ, Kreinick CJ. Brachial plexus palsy: An old problem revisited. Am J Obstet Gynecol 1992;166:1673–6.
3. Graham EM, Forouzan I, Morgan MA. A retrospective analysis of Erb's palsy cases and their relation to birth weight and trauma at delivery. J Matern Fetal Med 1997;6:1–5.
4. Ouzounian JG, Korst LM, Phelan JP. Permanent Erb palsy: A traction-related injury? Obstet Gynecol 1997;89:139–41.
5. Iffy L, Varadi V, Jakobovits A. Common intrapartum denominators of shoulder dystocia related birth injuries. Zentralbl Gynakol 1994;116:33–7.
6. Ecker JL, Greenberg JA, Norwitz ER, Nadel AS, Repke JT. Birth weight as a predictor of brachial plexus injury. Obstet Gynecol 1997;89:643–7.
7. Nocon JJ, McKenzie DK, Thomas LJ, Hansell RS. Shoulder dystocia: An analysis of risks and obstetric maneuvers. Am J Obstet Gynecol 1993;168:1732–7.
8. Morrison JC, Sanders JR, Magann EF, Wiser WL. The diagnosis and management of dystocia of the shoulder. Surg Gynecol Obstet 1992;175:515–22.
9. Walle T, Hartikainen-Sorri AL. Obstetric shoulder injury. Associated risk factors, prediction and prognosis. Acta Obstet Gynecol Scand 1993;72:450–4.
10. McFarland LV, Raskin M, Daling JR, Benedetti TJ. Erb/Duchenne's palsy: A consequence of fetal macrosomia and method of delivery. Obstet Gynecol 1986;68:784–8.
11. Hardy AE. Birth injuries of the brachial plexus. Incidence and prognosis. J Bone Joint Surg Br 1981;63:98–101.
12. Levine MG, Holroyde J, Woods JR, Siddiqi TA, Scott M, Miodovnik M. Birth trauma: Incidence and predisposing factors. Obstet Gynecol 1984;63:792–5.
13. Gordon M, Rich H, Deutschberger J, Green M. The immediate and long-term outcome of obstetric birth trauma. I. Brachial plexus paralysis. Am J Obstet Gynecol 1973;117:51–6.
14. Adler JB, Patterson RL Jr. Erb's Palsy. Long term results of treatment in eighty-eight cases. J Bone Joint Surg Am 1967;49:1052–64.
15. World Health Organization. International Classification of Diseases. 9th rev. Geneva: World Health Organization, 1997.
16. Hankins GD, Clark SL. Brachial plexus palsy involving the posterior shoulder at spontaneous vaginal delivery. Am J Perinatol 1995;12:44–5.
17. Dunn DW, Engle WA. Brachial plexus palsy: Intrauterine onset. Pediatr Neurol 1985;1:367–9.
18. Scheller JM, Nelson KB. Does cesarean delivery prevent cerebral palsy or other neurologic problems of childhood? Obstet Gynecol 1994;83:624–30.
19. Rouse DJ, Owen J, Goldenberg RL, Cliver SP. The effectiveness and costs of elective cesarean delivery for fetal macrosomia diagnosed by ultrasound. JAMA 1996;276:1480–6.
20. Perlow JH, Wigton T, Hart J, Strassner HT, Nageotte MP, Wolk BM. Birth trauma. A five-year review of incidence and associated perinatal factors. J Reprod Med 1996;41:754–60.
21. Specht EE. Brachial plexus palsy in the newborn. Incidence and prognosis. Clin Orthop 1975;110:32–4.
22. al-Qattan MM, al-Kharfy TM. Obstetric brachial plexus injury in subsequent deliveries. Ann Plas Surg 1996;37:545–8.
23. Ubachs JMH, Slooff ACJ, Peeters LLH. Obstetric antecedents of surgically treated obstetric brachial plexus injuries. Br J Obstet Gynaecol 1995;102:813–7.