Obstetrics & Gynecology 2001;98:1038-1040
© 2001 by The American College of Obstetricians and Gynecologists
Is Routine Hemoglobin and Hematocrit Testing on Admission to Labor and Delivery Needed?
Gordon B. Sherard, III, MD and
Edward R. Newton, MD
From the Department of Obstetrics and Gynecology, East Carolina University-Brody School of Medicine, Greenville, North Carolina.
Address reprint requests to: Edward R. Newton, MD, Department of Obstetrics and Gynecology, East Carolina University-Brody School of Medicine, 600 Moye Blvd., Greenville, NC 27858; E-mail: newtoned{at}mail.ecu.edu.
 |
ABSTRACT
|
|---|
OBJECTIVE: Testing hemoglobin and hematocrit values in labor and delivery at term is a routine practice at many centers. This necessity was evaluated by comparing values obtained at 26–28 weeks versus those obtained at term.
MATERIALS: This is a prospective, observational study. All patients at least 37 weeks presenting to labor and delivery were evaluated during the study period. Enrollment criteria included having either a hemoglobin or hematocrit between 26 and 28 weeks and again on admission with term labor. Deliveries at less than 37 weeks and pregnancies complicated by multiple gestations, hemoglobinopathies, and hypertensive disorders were excluded. The World Health Organization (WHO) definition of anemia was used as well as an operationally defined threshold, below which local practitioners would change management (hemoglobin 8 g/dL, hematocrit 25%). The compliance of patients to any form of iron therapy was evaluated by questioning patients on admission.
RESULTS: One hundred and one patients met enrollment criteria. At 26–28 weeks, 20 patients were anemic by WHO criteria. On admission to labor and delivery at term, relative to 26–28-week values, hemoglobin had increased from 11.1 g/dL to 11.6 g/dL (P < .01), and hematocrit increased from 31.5–34.3% (P < .01). Eleven patients had term values that were lower than 26–28-week values. Of these, five patients met WHO criteria at 26–28 weeks. However, no value at term was below the operationally defined value.
CONCLUSION: The frequency of anemia fell from 20% at 26–28 weeks to 11% at term. The mild anemia at term did not change local management in any patient. Thus, if the value obtained at 26–28 weeks is acceptable (non-anemic by WHO criteria), the routine testing of these values at term can be avoided, resulting in significant cost savings.
Routine laboratory screening is standard procedure for most pregnant women admitted at term to a labor and delivery unit. Previous studies have questioned the value of routine type and screen testing for expected vaginal deliveries and routine 36-week hemoglobin and hematocrit testing.1,2 Physiologic changes in plasma volume and red cell mass result in the nadir for hemoglobin and hematocrit concentration at 26–28 weeks’ gestation. A hemoglobin and hematocrit test is conveniently obtained at the time of the 50-g glucola screen, and, if anemia is detected, there is adequate time for effective therapy before delivery. In an attempt to enhance the cost-effectiveness of patient care, this study compares the values obtained between 26–28 weeks’ gestation with the values during term labor. Our hypothesis is that, in a low-risk patient, a normal hemoglobin and hematocrit at 26–28 weeks will predict a normal value at term.
|
MATERIALS AND METHODS
|
|---|
The University and Medical Center Institutional Review Board approved the study. The first 101 consecutive eligible patients had their postpartum charts reviewed for 26–28-week hemoglobin and/or hematocrit as well as for the hemoglobin and/or hematocrit on admission to the labor and delivery unit at Pitt County Memorial Hospital from July 1, 1999, to August 31, 1999. Study criteria included having a hemoglobin or hematocrit between 26 and 28 weeks’ gestation and a repeat hemoglobin or hematocrit on admission to labor and delivery at term. Preterm deliveries (less than 37 weeks) and pregnancies complicated by multiple gestations, hemoglobinopathies, and hypertensive disorders were excluded.
The primary outcome was the frequency of labor and delivery hemoglobin and/or hematocrit test results below a critical threshold relative to a 26–28-week hemoglobin and/or hematocrit result in the same patient. The World Health Organization (WHO) definition (hemoglobin of 10 g/dL and hematocrit of 30%) was used for comparison across studies. We operationally define a more severe threshold (hemoglobin of 8 g/dL and hematocrit of 25%) because most physicians do not transfuse blood for mild anemia and instead would change a type and screen to a type and cross. All local obstetric specialists (n = 31) determined the operational definition. Each was surveyed for when a labor hemoglobin and/or hematocrit would change management. The local survey responses for the operationally defined values are shown in Table 1
.
View this table:
[in this window]
[in a new window]
|
Table 1. Operational Threshold for Change in Management. At What Value Would You Change a Type and Screen to a Type and Cross?
|
|
The data were analyzed using paired t tests for continuous data with normal distributions, Mann-Whitney U test for non-normal distribution, 
2 tests for categoric data, and Fisher exact test for cell counts less than 5. A P value < .05 was considered significant. The statistical package used was Number Cruncher Statistical System (NCSS)-Statistical analysis and data analysis software (Kaysville, UT).
The sample size was justified using the premise that the failure to identify as few as one patient with a "normal" hemoglobin and/or hematocrit at 26–28 weeks with a critically low hemoglobin and/or hematocrit in labor at term might not justify elimination of a labor hemoglobin and/or hematocrit. We choose a sample size to establish with a power of 0.97 (97%) that the probability (with a 95% confidence interval) of missing an adverse event (ie, a normal hemoglobin and/or hematocrit at 26–28 weeks with a critically low hemoglobin and/or hematocrit at term) was less than 5%.3 A sample size of 100 established this power and confidence.
|
RESULTS
|
|---|
During a 6-week period, 101 patients met admission criteria of the study and were enrolled. The demographics of the obstetric population are described in Table 2. Twenty-five percent of patients received their prenatal care at local health departments, with the remainder receiving prenatal care at the University clinic. Our population was a typical setting for a University population in the rural mid-Atlantic area.
All patients were prescribed prenatal vitamins, and 81% (81 of 101) of patients reported compliance with the vitamins. The generic prenatal vitamin available at our University clinic contains 65 mg of iron and 1 mg of folate. Patients (n = 20) were started on iron therapy (ferrous sulfate, 325 mg, once a day) if, at 26–28 weeks’ gestation, their hemoglobin and/or hematocrit values were below 10 g/dL or 30%, respectively. On admission at term, patient compliance with the prescribed iron was addressed and noted to be less than 50% (n = 9).
Table 3 describes the comparison of 26–28-week hemoglobin and hematocrits to testing at term (more than 37 weeks). At 26–28 weeks, 20 patients (20%) were anemic by WHO standards and none by our operational threshold.
On admission to the labor and delivery unit at term, hemoglobin increased a mean of 0.5 g/dL (P < .005) and hematocrit 2.8% (P < .001). Eleven percent (11 of 101) of patients had values at term that were less than the values at 26–28 weeks. Of these 11, only five had values classified as anemic by the WHO. Among 20 patients who were anemic at 26–28 weeks by WHO criteria, only four remained anemic at term. Despite the poor compliance (50%) with iron therapy, 80% of anemic patients at 26–28 weeks had become non-anemic at term, by WHO criteria.
No patient had term values less than the operationally defined hemoglobin of 8 g/dL and hematocrit of 25%. The range of hemoglobin values was 9.3–11.3 g/dL and hematocrit values was 27.6–33% at term among women whose labor and delivery values were less than the mid-trimester values. Among these patients, the average difference between the term values and the 26–28-week values in the hemoglobin was 0.72 g/dL and for the hematocrit, 3.6%.
|
DISCUSSION
|
|---|
In this study, we established that in low-risk populations a normal hemoglobin and/or hematocrit at 26–28 weeks predicts a normal hemoglobin and/or hematocrit at term in labor. Because these values increase in most women during the remainder of the pregnancy, the necessity for obtaining values in labor and delivery is questionable. If the value at 26–28 weeks is acceptable (non-anemic)—and in this study the averages were 11.1 g/dL for hemoglobin and 31.5% for hematocrit—testing upon admission to labor and delivery is not necessary. The occurrence of 11 instances where the at-term values were actually lower is somewhat surprising. However, half of these were above anemia levels, and the lowest hemoglobin was 9.3 g/dL. While meeting the criteria for anemia, this value could still be considered acceptable for an expected vaginal delivery by the operationally defined limit (hemoglobin over 8 g/dL).
The two previous similar studies have been performed looking at the cost-effectiveness of routine screening for anemia in the late third trimester and the usefulness of obtaining type and screens on admission for expected vaginal delivery. Neither addressed the issues of our current study. Lee and Patrissi2 examined 36-week hemoglobin and hematocrits compared with 28-week and labor admission values. They demonstrated significantly increasing values from 28 to 36 weeks to term labor. Ransom et al1 evaluated over 16,000 patients admitted for an expected vaginal delivery and found a 0.47% transfusion rate, eliminating the need for type and screen testing in this population.
Our study is applicable to the low-risk patient admitted for expected vaginal delivery. We did not evaluate twins, prematurity, hypertensive disorders, or hemoglobinopathies, and we do not recommend applying our study results to those populations at this time. It might be appropriate for future studies to evaluate twins, prematurity, and hypertensive disorders, but we would counsel against any such study involving patients with hemoglobinopathies.
The advantages of limiting routine hemoglobin and hematocrit testing in labor are a reduction in discomfort and cost. A hemoglobin and hematocrit at 26–28 weeks can be obtained at routine glucola screening. A separate blood draw in labor might be avoided in low-risk patients. The cost in our hospital’s laboratory for a hemoglobin and hematocrit is $11 apiece, $22 for both. If a complete blood count were substituted for the hemoglobin and hematocrit, the cost would almost be doubled to $39. If a routine complete blood count was omitted from the uncomplicated, expected vaginal delivery, the potential cost savings at our institution (around 2500 term deliveries) would be $110,000 per annum. If these same principles were applied nationwide, the savings could be over $100 million per annum (based on 3.9 million deliveries in 1999). Whereas cost-savings is easy to judge, another important issue that is more difficult to appreciate is the benefits of decreasing unnecessary work for nursing, patient-care teams, and laboratory personnel. Manpower could be used more effectively, and the quality of care could be improved by eliminating unnecessary expenses.
All patients at the authors’ institution have routine hemoglobin and hematocrit testing on the first postpartum day that serves as a "safety net" to capture any patients with unexpectedly low postpartum laboratory values. The efficacy of routine postpartum hemoglobin and/or hematocrit in a low-risk population needs study. Perhaps the routine use of postpartum prenatal vitamins would eliminate the need for routine postpartum testing in patients with normal blood loss during the delivery process.
The authors also acknowledge that our sample size of women with anemia at 26–28 weeks (WHO criteria) is small (n = 20). However, none had sufficiently severe anemia in labor that would change management in the majority of our providers. Iron therapy remains appropriate for these patients, and compliance is a major issue. Until a larger sample size is examined, we recommend a labor evaluation of hemoglobin and hematocrit be performed in women who are anemic at 26–28 weeks.
|
Footnotes
|
|---|
PII S0029-7844(01)01586-1
Received February 6, 2001.
Received in revised form July 3, 2001.
Accepted July 19, 2001.
|
REFERENCES
|
|---|
1. Ransom SB, Fundaro G, Dombrowski MP. The cost-effectiveness of routine type and screen admission testing for expected vaginal delivery. Obstet Gynecol 1998;92:493–5.[Abstract]
2. Lee D, Patrissi GA. Routine 36-week hemoglobin and hematocrits: Are they necessary? Mil Med 1994;159:201–3.[Medline]
3. Hanley JA, Lippman-Hand A. If nothing goes wrong, is everything all right? Interpreting zero numerators. JAMA 1983;249:1743–5.[Medline]
TOP
ABSTRACT
MATERIALS AND METHODS
