Dataset

This was an observational cohort study of nonanomalous term singleton deliveries where complete and validated cord blood gas analyses were recorded.

Details of all deliveries at a major teaching hospital, including cord gas analysis if performed, were prospectively recorded in a database. Cord gases are taken at approximately half of all deliveries. These are manually entered after birth: so to avoid potential transposition errors, we merged outcome data, using hospital number and date, with the output from the unit blood gas analyzer.

Of appropriate deliveries from June 23, 2005, through Dec. 31, 2009, where cord samples were taken, we excluded all where only 1 vessel was sampled or any values, eg, bases deficit, pH, or partial pressure of carbon dioxide (pCO2), were missing. We then excluded unphysiological values, ie, where the arteriovenous (AV) pCO2 or hydrogen ion (H+) difference was negative. We also excluded unreliable results where samples might be contaminated, or the same vessel sampled twice, with an adaptation of Westgate and Greene. We excluded cases in the lowest 5% for both pCO2 AV difference (≤2.5000) and H+ ion AV difference (≤0.0014). We used hydrogen ion difference, since excluding by pH difference preferentially excludes those results with a low pH and a narrow AV difference because pH is a logarithmic number. Finally we limited our analysis to moderately acidemic (ApH <7.1 and ≥7.0) and severely acidemic (ApH <7.00) cohorts. The Figure shows a summary of exclusions.

Summary of exclusions

Knutzen. Base deficit in acidemic term neonates. Am J Obstet Gynecol 2015 .

As the object of this study was simply to examine the relationship between ApH and ABD as risk factors for adverse outcomes, we did not confine analysis to babies delivered after labor, or examine other potential risk factors for adverse outcomes, such as birthweight.

Cord gas analysis

Unit policy is that paired cord gases should be taken if there has been electronic fetal monitoring or meconium, or at midwife or doctor discretion, eg, low Apgar scores. The sampling rate for the period was 45%. The cord is double-clamped immediately after delivery at a minimum length of 10 cm with the placenta in situ. Arterial and venous samples are taken in preheparinized labeled syringes and processed within 15 minutes in a Radiometer ABL800 blood gas analyzer (Radiometer Medical ApS, Brønshøj, Denmark). Radiometer reports base excess (BE) in 2 compartments: blood and extracellular. As the more physiological value, arterial extracellular fluid BE (ABE[Ecf]) was the measure used. Unfortunately the values of ABE(Ecf) were not calculated by the analyzer for the year 2005. For consistency of language the sign was changed and base deficit values used. For the entire cohort the values of base deficit (Ecf) were calculated from pH and pCO2. Firstly bicarbonate (HCO ₃ ) was calculated using the Henderson-Hasselbalch equations (HCO ₃ = 0.03 × pCO ₂ × 10 ^[pH–6.1] ). These values were then entered into the algorithm used by Westgate and Greene to calculate base deficit (Ecf).

Outcome measures

Data on neonatal events were initially assessed using International Statistical Classification of Diseases codes of all babies. Neonatal records, discharge summaries, and imaging reports of babies admitted to the neonatal unit prior to maternal discharge were inspected.

We selected outcome measures in conjunction with a neonatologist to reflect clinical condition at birth, neurological involvement, and/or multisystem involvement. We used encephalopathy (grade 2 and 3) and/or death as well as more frequent outcomes of neonatal unit admission and Apgar score <7 at 5 minutes as the main outcome measures. We also developed composite outcomes to identify babies with multisystem involvement. The composite neurological outcome included those with ≥1 of: encephalopathy of any grade, seizures, abnormal movement that was not classified as seizures, abnormal tone, and abnormal monitoring or imaging (≥1 of: cerebral functioning monitoring, cranial ultrasound, electroencephalogram, or magnetic resonance imaging are abnormal). The composite systemic outcome included those with ≥1 of: ventilated for >24 hours, hepatic impairment (alanine amniontransferase ≥100 IU/L in first 48 hours), renal impairment defined as a creatinine of ≥80 micromol/L in the first 48 hours, coagulopathy first 24 hours, and need for inotropes. Neurological and systemic composite outcomes were combined to identify those with at least 1 neurological outcome and at least 1 other system involved.

Institutional review board approval was granted in September 2005 (05/Q1605/110) and updated on Feb. 16, 2010.

Statistical analysis

Analysis was carried out using SPSS 20.0 (IBM Corp, Armonk, NY). The distribution of ApH values was skewed, so medians and Mann-Whitney U tests were used to examine the relationship among ApH, ABD(Ecf), and each adverse outcome within the acidemic cohorts.

Hierarchical logistic regressions were used to determine whether, in acidemic neonates, ABD(Ecf) made a significant contribution to predicting adverse outcomes in addition to the predictive value of ApH. ApH, then ABD(Ecf), and then their cross-product (a product of standardized ApH and ABD[Ecf]) were entered into the regression equations in 3 steps. The cross-product was included to investigate whether ApH and ABD interact (ie, their effect together is greater than the sum of their parts). We examined the change in the model fit (the accuracy with which the model predicts the outcome) with the addition of each predictor.

We also looked at the statistics for the final model when all 3 predictors were in the regression equation. The statistics for each predictor in the final model quantify its effect on the outcome variable when controlling for the other predictors in the model.

As ApH is on a logarithmic scale, the regression analysis produces extremely large odds ratio estimates. To compensate for this, ApH and ABD(Ecf) measures were standardized (expressed as the number of SD from the mean) before entry into regression, which is an accepted practice when dealing with variables distributed on very different scales.

Cohort characteristics

There were 520 (5.9%) neonates in the cohort who were born with an ApH of <7.1. Of these 84 (0.95%) had an ApH <7.0. In all, 436 neonates had ApH <7.1 and ≥7.0. Base deficit and ApH were significantly negatively correlated (r = –0.51, P < .001), demonstrating the association of greater base deficit with lower pH. Within the cohort there were 7 sets of incomplete neonatal records, resulting in those babies having incomplete data. Table 1 summarizes the characteristics of the cohort.

Neonatal outcomes

Among all acidemic neonates, 2.3% developed encephalopathy and/or died. This figure was higher (8.5%) among the severely acidemic subgroup. Detailed incidences of neonatal adverse outcomes are summarized in Table 2 .

Prediction of adverse outcomes

Among all the acidemic neonates (ApH <7.1) both ApH and ABD(Ecf) were significantly related to each adverse outcome. Specifically, both measures were significantly lower on average among those affected than those who were unaffected ( Table 3 ). In the smaller pH <7.0 group fewer of these differences reached statistical significance.

The results of hierarchical regression are shown in Table 4 . For each outcome variable the model with only ApH (the first step) significantly predicted the outcome, and the addition of ABD(Ecf) or the cross-product did not improve the prediction. Looking at the final model statistics, where all 3 variables are entered, ApH remained a significant predictor of each adverse outcome when controlling for ABD(Ecf) and the cross-product. Neither ABD(Ecf) nor the cross-product were significant predictors of outcome when controlling for the other two.

Table 4

Relationship of arterial pH and arterial base deficit in predicting adverse outcomes in whole acidemic cohort (neonates with arterial pH <7.1)

Adverse outcome	Predictors	Model change	P value	Final model	P value
		χ 2		Wald χ 2
Encephalopathy grade 2/3 and/or death	1. ApH	17.3	< .001	4.4	.036
	2. ABE(Ecf)	0.03	.868	0.2	.689
	3. Cross-product	1.2	.283	1.1	.286
Neonatal unit admission	1. ApH	64.0	< .001	24.4	< .001
	2. ABE(Ecf)	0.02	.893	0.002	.965
	3. Cross-product	0.7	.409	0.7	.417
Apgar <7 at 5 min	1. ApH	43.6	< .001	20.2	< .001
	2. ABE(Ecf)	0.001	.972	0.001	.974
	3. Cross-product	0.000007	.998	0.000007	.998
Composite neurological adverse outcome	1. ApH	36.3	< .001	17.8	< .001
	2. ABE(Ecf)	0.007	.934	0.00002	.997
	3. Cross-product	0.02	.885	0.02	.885
Systemic involvement	1. ApH	35.6	< .001	17.7	< .001
	2. ABE(Ecf)	0.03	.857	0.08	.781
	3. Cross-product	0.1	.757	0.1	.756
Composite neurological adverse outcome and systemic involvement	1. ApH	43.2	< .001	13.6	< .001
	2. ABE(Ecf)	0.2	.676	0.003	.958
	3. Cross-product	0.7	.403	0.7	.407

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

1. Predictors of 30-day readmission and impact of same-day discharge in laparoscopic hysterectomy
2. Maternal and fetal morbidity associated with uterine rupture of the unscarred uterus
3. Does magnesium exposure affect neonatal resuscitation?
4. Additional uses of a vaginal fornices delineator

Stay updated, free articles. Join our Telegram channel

Join