It is highly controversial whether exposure to anesthesia and surgery during pregnancy places either the mother or fetus at additional risk for complications.1 There appears to be an increased incidence of spontaneous abortion in women undergoing surgery during the first trimester of pregnancy.2,3 There may also be an increased risk of congenital defects in the offspring of women exposed to anesthesia and surgery during this period.4,5 Recognizing a pregnancy allows a woman and her physician to make an informed decision before proceeding with elective surgery. Contentious scientific evidence aside, it is clear that elective surgery is postponed when an unanticipated pregnancy is discovered immediately before the procedure.6,7 Identifying pregnancy during this time alters clinical management.
Manley et al. prospectively studied a large cohort of 2056 patients to determine the incidence of unrecognized pregnancy in women presenting for elective ambulatory surgery and the impact of a positive pregnancy test in this population.6 One criticism of their study was that the test was performed as many as 6 days before the scheduled surgery. Human chorionic gonadotropin (hCG) concentrations can increase from undetectable to detectable in this period. Testing on the day of surgery would eliminate this problem and identify as many pregnancies as possible.8
The American Society of Anesthesiologists (ASA) House of Delegates approved the report of the Task Force on Preoperative Evaluation, which stated:
The Task Force recognizes that a history and physical examination may be insufficient for identification of early pregnancy. Pregnancy testing may be considered for all female patients of childbearing age. Clinical characteristics to consider include an uncertain pregnancy history or a history suggestive of current pregnancy.9
The ASA effectively gives individual physicians and hospitals the opportunity to set their own policies and practices relating to preoperative pregnancy testing. To optimize informed decision making, the Hospital for Special Surgery implemented a pregnancy testing policy in November 2004 in which women of childbearing age were routinely tested for the presence of hCG in urine on the day of surgery. We report the results and associated costs of this policy over 1 yr and hypothesize that it is effective in detecting otherwise unrecognized pregnancy, subsequently altering the surgical course for elective surgery.
Additionally, the report of the Task Force on Preoperative Evaluation stated:
The current literature is not sufficiently rigorous to permit an unambiguous assessment of the clinical benefits or harms associated with selected preoperative test findings.9
We believe that this study adds to the current body of evidence, as no study has reported the incidence of positive pregnancy when testing is performed on the day of surgery in patients undergoing elective orthopedic procedures.
After IRB approval, the medical records of all surgical patients between January 3, 2005, and January 2, 2006, were retrospectively reviewed. As this was a retrospective review of an existing policy, the IRB issued a waiver of informed consent as pursuant to the Code of Federal Regulations Title 45, Part 46.116(d).
As directed by the policy, all women of childbearing age (defined as the age between initial reported menses, and 1 yr after last reported menses) would undergo urine testing for hCG on the day of surgery. During the preoperative telephone call made by nursing on the day before surgery, female patients would be informed of the test and the option of refusal. On the day of surgery at the time of admission, holding area nurses assessed all patients for testing eligibility. Eligible patients were then asked to void and submit the sample for testing. Patients could refuse the test after being advised of the associated risks. This decision would be discussed among the patient, surgeon, and anesthesiologist and if any individual was uncomfortable with proceeding, the case would be cancelled.
The following data were collected retrospectively using documentation by holding area nurses: eligibility for pregnancy testing as outlined in the policy, reasons for exclusion from testing (patient report of hysterectomy or bilateral salpingo-oophorectomy, but not inclusive of tubal ligation), and patient refusal. The reported use of any contraceptive device or medication was not considered a criterion for exclusion from testing. The results of the urine hCG were also recorded, as was the disposition of all patients with a positive result.
The pregnancy test used for both urine and serum samples was the ICON®II ImmunoConcentration™ Assay (Beckman Coulter, Fullerton, CA), which incorporates monoclonal antibody technology. After exposure to the test reagents, specimens produce a circular blue spot on the assay membrane if they contain hCG. Intensity approximating or more than the reference is reported as “positive.” If the intensity is less than that of the positive reference, it is reported as “weak positive.” In evaluations reported by the manufacturer in the package insert, urine hCG concentrations ≥20 mIU/mL and serum hCG concentrations ≥10 mIU/mL were detected by an occurrence of any blue spot 100% of the time. When tested with serum and urine samples known to lack hCG, the assay was found to be 100% specific.
As is consistent with hospital policy, any positive result was discussed with the patient by an attending physician. After this discussion, the positive urine test result was confirmed with a serum hCG test with the patient’s consent. Patients with a negative serum test proceeded with surgery as scheduled, whereas those with a positive serum hCG were referred to their obstetrician for consultation and follow-up care.
The cost of implementing this policy during the year was estimated using the cost of the reagents and laboratory technician time as supplied by the Chief Laboratory Technologist. Information of laboratory technician compensation was provided by the Director of Human Resources. The details of this calculation are given in the Results section. Additionally, a number-needed-to-treat analysis was performed.10
The positive predictive value for each positive designation was calculated as the ratio of true positives to all positives (true plus false). Since no additional confirmatory testing was performed to determine if any patients with a negative urine test result was truly pregnant at the time of testing (false negatives), a negative predictive value calculation could not be performed.
From January 3, 2005, to January 2, 2006, 8610 women had elective surgery at our institution, of whom 2595 met the criteria for urine pregnancy testing (Fig. 1). Two women refused testing, and a third woman was known to be 12 wk pregnant. After documentation in the medical record, these patients underwent the scheduled procedures. Among the women with a positive hCG, one had a reported dilation and extraction performed 2 days before surgery. Since one would expect hCG levels to be increased, this positive test was excluded from further analysis. Of the remaining 2588 patients, 8 had positive urine test results (3 weak positive and 5 positive) and 7 went on to have serum hCG testing. A summary of these patients and testing outcomes are described in Table 1.
Based on the analyzed data in our population, the incidence of unrecognized true positive pregnancies on the day of surgery is 4 of 2588, or 0.15% (95% CI: 0.003-0.31), whereas the incidence of true plus false positives is 8 of 2588, or 0.30% (95% CI: 0.10-0.52). The positive predictive value of any urine hCG positive designation (positive and weak positive) is 50%. The positive predictive value increases to 80% for positive designation only. Furthermore, the positive predictive value reaches 100% when a positive urine result is combined with a positive serum result. The number-needed-to-treat to detect one true positive result is 647 (95% CI: 327-661).
The cost of each ICON®II ImmunoConcentration™ Assay kit at our institution is $2.50. According to the Chief Laboratory Technologist, a urine or serum hCG assay takes approximately 7 min to perform, of which 4 min is labor-intensive when the laboratory technician must actively attend to the test duties. The remaining 3 min are “wait time” when the laboratory technician is able to perform other tasks. Thus, we assume that the technician time associated with each assay is 4 min. Since 2592 urine pregnancy tests were performed in 2006, the total labor was 172.8 h. The average technologist’s compensation package corresponds to $38/h, resulting in $6566 in labor. Therefore, the labor and material costs for 1 yr of urine hCG tests is $13,046, or $3262 per true positive and $5.03 per urine test. Including the 9 serum hCG tests that were performed, the cost becomes $3273 per true positive.
The ASA Task Force on preoperative pregnancy testing developed a policy addressing unrecognized pregnancy to optimize patient safety, improve risk management, and improve informed decision making. There are several reasons this policy was instituted. Up to 15% of known pregnancies miscarry before 20 wk and up to 50% of unrecognized pregnancies miscarry during the first trimester.11 Studies suggest that there may be an increased risk of spontaneous abortion associated with surgery and anesthesia. Duncan et al. found an increased risk of spontaneous abortion associated with surgery under general anesthesia in the first or second trimester.3 Similarly, in a retrospective survey, Brodsky et al. found that surgery with anesthesia during the first or second trimester was associated with an increased incidence of spontaneous abortion, and Mazze and Kallen found an increased incidence of very low and low birth weight infants.2,12 Kallen and Mazze found an association between operations during the first trimester and congenital anomalies, specifically neural tube defects.4 Sylvester et al. found an association between hydrocephalus with other defects in offspring of mothers exposed to anesthesia and surgery in the first trimester.5
Although the reports described above are conflicting as to cause and effect, should a spontaneous abortion occur after surgery, or the baby be born with a congenital anomaly, this may be attributed to the surgery or anesthetic. Avoiding such a scenario is in the best interests of the patient, doctor, and hospital administration. In our orthopedic hospital, most patients who undergo surgery are exposed to ionizing radiation, either from pre- or postoperative radiographs, or intraoperative fluoroscopy. This should be avoided during pregnancy, particularly in the first trimester, because of the known increased incidence of congenital anomalies.13
In the first full year of testing, two women with positive pregnancy tests resulting in surgery cancellations went on to have spontaneous abortions. Had testing not been performed, they may have blamed their loss on the surgery, resulting in possible culpability at the hands of the health care providers involved. A third patient whose pregnancy was discovered before surgery elected to have the pregnancy terminated. The testing policy allowed her to make this difficult decision earlier in the pregnancy and without the added complexity of a recent exposure to surgery and anesthesia. The fourth unexpectedly positive test was due to an ectopic pregnancy in a patient who was status post a tubal ligation. This prevented potential morbidity complicating the perioperative period of elective orthopedic surgery and also allowed the ectopic pregnancy to be treated before it became symptomatic, or ruptured.
Surgery was cancelled unnecessarily in only two instances, one being soon after a termination of pregnancy. Both of these patients underwent surgery approximately 1 wk later, after consultation with an obstetrician-gynecologist. Three other patients, all perimenopausal, were inconvenienced with a weak positive urine hCG result. The absence of pregnancy was supported by a negative serum hCG, and surgery proceeded in all three cases.
It is of interest that of the four women with false positive test results, three reported more than 1 yr had elapsed since their last menstrual period. This is an oversight of the policy, as preoperative nurses occasionally send urine for testing before the patient’s complete history is obtained. If testing had been limited to those women meeting policy criteria, there would have been only one false positive (Table 1, Patient #7).
Two patients were excluded from data analysis. One was known to be pregnant and was therefore not tested. The other had a positive pregnancy test after experiencing a termination of pregnancy 2 days before surgery. The purpose of the policy, and of interest to clinicians and patients, is to identify unrecognized pregnancies. If pregnancy status is known, testing is unnecessary.
Although we report an incidence of positive tests of 0.15%, the reported incidence at other institutions when all women are screened has ranged between 0.3 and 2.2%.6,14,15 This difference in other populations is likely related to the variation in socioeconomic status and the extent of patients’ involvement in the medical system. There could also be a difference in the likelihood of intentionally withholding a known pregnancy from caregivers during elective surgery. One must take these factors into account when considering the applicability of these reported results to other health-care institutions.
The cost of finding these three pregnancies and one ectopic pregnancy is not insignificant. We estimated the yearly cost to be roughly $13,050, which represents only the expenditures of testing materials and labor of performing the urine and serum assays in the laboratory. This does not include the personnel costs in the preoperative area for obtaining specimens and transporting them to the laboratory. The existing personnel absorbed this task, i.e., no new employees were hired to implement the pregnancy testing policy. We have no way of quantifying operating room delays that may have resulted from this testing policy. Now that urine pregnancy testing has become an established part of the preoperative routine, operating room delays are rare. Before implementing this policy, patients were scheduled to arrive at the hospital 2 h preoperatively. In facilities using less preparation time, a higher incidence of operating room delays is possible.
There are several limitations to the study, mostly related to its retrospective design. There may have been documentation errors by holding area nurses. The pregnancy testing policy, as stated in the methods section, is dependant upon patients self-reporting their menstrual history. Policies using different testing criteria could have different results. This study was not designed to compare the reliability of testing on the day of surgery to an earlier testing day. It is notable that Patient #3 in Table 1 reported her last menstrual period to be 3 days before testing, suggesting that an earlier test may have been negative.
It seems that, based on our experience and most other reports, an occasional pregnant woman will undergo surgery during the first trimester in the absence of mandatory testing. The known risks of spontaneous abortion and radiation exposure, as well as possible congenital anomalies, must be weighed against the costs of screening. Additionally, screening may decrease litigation, although potential cost savings are difficult to quantify. The costs associated with pregnancy testing are justified by the potential risk to mother and fetus; however, each institution must make its own calculated judgment before implementing its own policy.
In summary, a mandatory pregnancy testing policy using urine hCG measurements was successfully implemented at the Hospital for Special Surgery. During its first full year, three pregnancies and one ectopic pregnancy were detected, whereas four women had false positive results. The approximate cost was $3273 for each pregnancy. A future prospective study addressing patient satisfaction issues, specific consent issues, and implementation of a pregnancy testing policy in other patient populations would be intriguing.
1. Kuczkowski KM. Nonobstetric surgery during pregnancy: what are the risks of anesthesia? Obstet Gynecol Surv 2004;59:52–6
2. Brodsky JB, Cohen EN, Brown BW Jr, Wu ML, Whitcher C. Surgery during pregnancy and fetal outcome. Am J Obstet Gynecol 1980;138:1165–7
3. Duncan PG, Pope WD, Cohen MM, Greer N. Fetal risk of anesthesia and surgery during pregnancy. Anesthesiology 1986;64:790–4
4. Kallen B, Mazze RI. Neural tube defects and first trimester operations. Teratology 1990;41:717–20
5. Sylvester GC, Khoury MJ, Lu X, Erickson JD. First-trimester anesthesia exposure and the risk of central nervous system defects: a population-based case-control study. Am J Public Health 1994;84:1757–60
6. Manley S, de Kelaita G, Joseph NJ, Salem MR, Heyman HJ. Preoperative pregnancy testing in ambulatory surgery. Incidence and impact of positive results. Anesthesiology 1995;83:690–3
7. Hennrikus WL, Shaw BA, Gerardi JA. Prevalence of positive preoperative pregnancy testing in teenagers scheduled for orthopedic surgery. J Pediatr Orthop 2001;21:677–9
8. Rosenberg MK. Preoperative pregnancy testing in ambulatory surgery. Anesthesiology 1996;84:1260
9. American Society of Anesthesiologists Task Force on Pre Anesthesia Evaluation. Practice advisory for pre anesthesia evaluation: a report by the American society of anesthesiologists task force on pre anesthesia evaluation. Anesthesiology 2002;96:485–96
10. Cook RJ, Sackett DL. The number needed to treat: a clinically useful measure of treatment effect. BMJ 1995;310:452–4
11. Wilcox AJ, Weinberg CR, O’Connor JF, Baird DD, Schlatterer JP, Canfield RE, Armstrong EG, Nisula BC. Incidence of early loss of pregnancy. N Engl J Med 1988;319:189–94
12. Mazze RI, Kallen B. Reproductive outcome after anesthesia and operation during pregnancy: a registry study of 5405 cases. Am J Obstet Gynecol 1989;161:1178–85
13. Swartz HM, Reichling BA. Hazards of radiation exposure for pregnant women. JAMA 1978;239:1907–8
14. Twersky RS, Singleton G. Preoperative pregnancy testing: “justice and testing for all.” Anesth Analg 1996;83:438–9
15. Wheeler M, Cote CJ. Preoperative pregnancy testing in a tertiary care children’s hospital: a medico-legal conundrum. J Clin Anesth 1999;11:56–63