Admission test as precursor of perinatal outcome: a prospective study

Arch Gynecol Obstet (2010) 282:377–382
DOI 10.1007/s00404-010-1406-4
MATERNO-FETAL MEDICINE

Admission test as precursor of perinatal outcome: a prospective study

Shivani Khandelwal · Mitra Dhanaraj · Atul Khandelwal

Received: 17 December 2009 / Accepted: 9 February 2010 / Published online: 9 March 2010© Springer-Verlag 2010

AbstractPurpose To test the reliability of the admission test toidentify the compromised fetus and thus reduce the neona-tal morbidity and mortality by early intervention.Methods A prospective analysis over a period of 1 yearfrom December 2007 to December 2008 included 100 ante-partum patients and were evaluated for perinatal outcomein two groups.Results In both low and high risk groups the incidence ofmeconium staining was 25 and 37.5% in patients with non-reactive traces as compared to 8.6 and 9.5%, respectively,in reactive traces with speciWcity of 90.9%. Clinicallydetected fetal distress was more common in patients withnonreactive test. Operative interference for fetal distresswas more in patients with nonreactive test. Occurence oflow Apgar score was more in patients with nonreactive test.Admission to neonatal unit was more in nonreactive thanreactive traces. Incidence of neonatal death was more innonreactive test. Incidence of low birth weight was more innonreactive trace group and more so in high risk group thanin low risk group.Conclusion Admission test may be best recommended inall patients irrespective whether they are in low or high risk

as incidence of neonatal morbidity is high 33.3% babiesrequired NICU admission and 33% babies expired in non-reactive tracing, in centers where advance facilities are notavailable. Whenever there is a nonreactive tracing furthertest should be carried out.

Keywords Admission test (AT) · Low risk pregnancy · High risk pregnancy · Perinatal outcome · Fetal heart rate (FHR)

Introduction

The goal of intrapartum monitoring is to detect maternal aswell as fetal complication. The admission test (AT) is ashort recording of the fetal heart rate (FHR) immediatelyafter admission to the labor ward. It is useful in detectingfetal distress already present at admission and also of somepredictive value for fetal well-being in the next few hoursof labor. Other investigations which have been studied asAT are vibro acoustic stimulation test, Doppler velocimetryand amniotic Xuid index. The use of AT is thus, to pick up acompromised fetus early in labor by observing their behav-ior to the stress of early labor [1]. The basis of FHR moni-toring is, in a real sense, fetal brain monitoring. The fetalbrain responds to various stimuli such as chemoreceptors,baroreceptors, and direct eVect of metabolic changes withinthe brain itself and signals to the fetal heart to alter it on amoment to moment basis. The initial response to chronic orslowly developing hypoxia is to increase cardiac output andredistribute this to the brain and heart. The increase in car-diac output is achieved by an increase in heart rate. Thismay be followed by a reduction in heart rate variability dueto brain stem hypoxia. Continuing hypoxia produces myo-cardial damage and heart rate decelerations. Acute hypoxia,

S. Khandelwal · M. DhanarajDepartment of Gynecology and Obstetrics, Herbertpur Christian Hospital, Herbertpur, Dehradun, India

A. KhandelwalDepartment of Urology, Pt.B.D.Sharma PGIMS, Rohtak, Haryana, India

S. Khandelwal (&)House No. 477, Sec 14, Rohtak, Haryana 124001, Indiae-mail: [email protected]

123

378 Arch Gynecol Obstet (2010) 282:377–382

in contrast to chronic hypoxia, results in a decrease in FHRinitially produced by chemoreceptor mediated vagal stimu-lation but eventually by myocardial ischemia. Metaboli-cally progressive fetal hypoxia results Wrst in a respiratoryacidemia and secondly in a metabolic acidemia with tissueinjury. With these underlying concepts, FHR monitoring isthe commonest method employed to assess fetal well-beingin labor. Generally the variability of the FHR is describedas having two components “short-term” and “long-termvariability”. Short-term variability is the beat to beat irregu-larity in the FHR and caused by the diVerence in ratesbetween successive beats of the FHR. It is caused by thepush–pull eVect of sympathetic and parasympathetic nerveinput, but the vagus nerve has main role in aVecting vari-ability. Long-term variability is generally 3–5 cycles/min.An acceleration is a visually apparent abrupt increase in thefetal heart baseline. They represent an intact neurohor-monal and cardiovascular control mechanism linked to fetalbehavioral states. Decelerations may be early, late or vari-able. Head compression probably causes vagal nerve acti-vation due to dural stimulation that mediates heart ratedeceleration [2]. A late deceleration is a smooth, gradualsymmetrical decrease in fetal heart rate beginning at orafter the peak of the contraction and returning to baselineonly after the contraction has ended [3]. Variable decelera-tion of the fetal heart rate is deWned as the visually apparentabrupt decrease in rate, the onset of which varies with suc-cessive contractions. A true sinusoidal pattern may be asso-ciated with serious fetal anemia, fetal distress and umbilicalcord occlusion and also with administration of meperedine,morphine etc.

Materials and methods

This was a prospective study and conducted in HerbertpurChristian Hospital, Herbertpur, Dehradun, India, over aperiod of 1 year from December 2007 to December 2008. Atotal of 100 patients were studied during this period anddivided into 2 groups which were comparable to each other:high risk group (50 patients)— postdated pregnancy, GDM,anemia, and oligohydramnios, BOH, IUGR and Rh nega-tive; low risk group (50 patients) age, parity, gestationalage, socioeconomic status were noted in all the women. 100antepartum patients admitted in the labor room wereincluded in the study according to the following criteria:

Inclusion criteria: delivery at the Herbertpur ChristianHospital, >34 weeks of gestation, pregnancy inducedhypertension, gestational diabetes mellitus, intrauterinegrowth restriction, post term pregnancy.Exclusion criteria: patients in active labor, <34 weeks,twins, hydramnios, previous cesarean within 3 years.

An admission FHR tracing was performed by a cardio toco-dynamometer on all admitted patients in early labor. AnFHR tracing was obtained for 20 min and then categorizedas normal, suspicious or abnormal based on FIGO recom-mendations. Labor was managed by the labor ward teamaccording to established protocol. ArtiWcial rupture ofmembranes (ARM) was performed when the patient was inactive stage of labor and the color of amniotic Xuid at mem-brane rupture was noted. The patients then underwent inter-mittent auscultation or continuous fetal monitoring.Outcome variables

Maternal: operative delivery for fetal distress-Instru-mental/Cesarean section.Neonatal: Apgar < 5 at 1 min, Apgar < 7 at 5 min,meconium staining of liquor, admission to the neonatalintensive care unit, neonatal mortality, birth weight.

The observer knew only the results of CTG and color of theliquor and was blinded to all other test results to removebias. The ethics of such a trial was discussed and theapproval of our hospital ethics committee obtained.

FIGO’s criteria of FHR trace [4]

Reactive/normal: normal acceleration (greater than15 beats), for greater than 15 s in 20 min, Traces withno accelerations but normal baseline variability (10–25 beats/minute), Normal baseline rate with earlydeceleration but with accelerations.Equivocal: normal baseline with no rate accelerationsin 20 min and reduced baseline variability (5–10 beats/min), abnormal baseline rate (greater then or equal to‘160 beats/min) with no accelerations, variable decel-erations without ominous signsOminous: baseline variability of <5 beats/min andabnormal baseline rate, repeated late decelerations,repeated variable decelerations with any of the follow-ing ominous signs-durations greater than 60 s anddecelerations greater than 60 beats from the baselineFHR, rebound tachycardia, slow recovery, reduced var-iability between decelerations, late component.

Statistical Analysis of the test and the outcome was mea-sured by Pearson Chi square test, using SPSS software. A pvalue of <0.05 was considered statistically signiWcant.

Results

Maximum numbers of patients were seen in the thirddecade: 60–70% patients were in the middle socioeconomicstatus, 30–35% of the patients were nulliparous (primigrav-ida) in both the groups. Maximum number of patients were

123

Arch Gynecol Obstet (2010) 282:377–382 379

of post-term pregnancy (28%) followed by PIH who were24%, GDM, anemia, and oligohydramnios 10%, BOH 8%,IUGR 6% and Rh negative 4% in the present study.

In low risk group, 92% (46) patients had reactive tracingand 8% (4) patients had nonreactive tracing. In high riskpatients, 84% (42) had reactive tracing and 16% (8) hadnonreactive tracing. The nonreactive tracing was more inthe high risk than in the low risk group.

In low risk group, 8.6% of the paients with reactive AThad meconium stained liquor, whereas 25% of the patientswhose AT was nonreactive had meconium stained liquor.In the high risk group, 9.5% of the patients who had ATreactive had meconium stained liquor whereas 37.5% of thepatient had meconium stained liquor when the AT was non-reactive. So it is observed that occurrence of meconiumwas high in patients with nonreactive AT in both groups(p < 0.05) as shown in Fig. 1. Table 1 shows the statistics.

Out of 100 patients, clinical fetal distress was seen in18% (16 of 88) of the patients with reactive trace and66.6% (8 of 12) of thepatients with nonreactive trace. Inlow risk group out of 46 with reactive traces 13%, (6) hadfetal distress clinically. Out of 4 patients with nonreactivetrace, 50% (2) had clinical fetal distress. In high risk grouppatients out of 42 with reactive trace 23.8% (10) had clini-cal fetal distress and out of 8 patients with nonreactivetrace, 75% (6) had clinical fetal distress. Thus, clinicallydetected fetal distress was more common in patients withnonreactive AT in both low and high risk groups. p < 0.05as shown in Fig. 2. Table 2 shows the statistics.

It was observed that out of 100 patients, 17 of 88(19.3%) with reactive AT and 6 of 12 (50%) with nonreac-tive test had LSCS for fetal distress. In low risk group, outof 46 patients with reactive trace, 20 (43%) had normalvaginal delivery, 9 (19.5%) underwent LSCS for fetal dis-tress, 1 (2.17%) had forceps delivery for fetal distress and16 (34.7%) underwent LSCS for other indications (likeNPOL, CPD, etc.). Out of 4 patients with nonreactive trace,1 (25%) had normal vaginal delivery and 3 (75%) under-went LSCS for fetal distress. The indication for LSCS inthe low risk group with reactive test was more in non-fetaldistress indications such as NPOL, CPD, etc. In 42 highrisk group of patients with reactive trace, 17 (40.4%) deliv-ered vaginally, 7 (16.6%) had LSCS for fetal distress and18 (40.4%) had LSCS for other indications. Out of 8 highrisk patients with nonreactive trace 4 (50%) delivered vagi-nally, 3(37.5%) had LSCS for fetal distress and 1 (12.5%)had LSCS for other indications. Hence it is observed thatpatients with non reactive AT had more incidences ofLSCS due to fetal distress than the patients with reactivetest (p < 0.05). This is shown in Tables 3 and 4.

It is noted that out of 100 patients, 8 babies had Apgarscore (A/s) 1–6 at 1 min and 4 had A/s 1–6 at 5 min as well.Of 88 patients who had reactive AT, low A/s at 5 min wasseen in 1 neonate (1.1%) and of 12 nonreactive traces lowA/s was seen in 3 (25%) neonates. In the low risk group outof 46 patients 1 (2.1%) neonate had A/s 7–10 at 1 min, butall 46 (100%) neonates had A/s 7–10 at 5 min. Out of 4patients with nonreactive traces, 2 (50%) neonates had A/s1–6 and 2 (50%) had A/s 7–10; of the two babies with A/s

Fig. 1 Admission test in relation to meconium stained liqour

Table 1 Admission test in relation to meconium stained liqour

Low risk (%)

High risk (%)

Overall (%)

Sensitivity 20 42.80 33.3

SpeciWcity 93.30 88.30 90.9

Positive predictive value 25 37.50 66.6

Negative predictive value 91.30 90.40 90.9

Fig. 2 Admission test in relation to fetal distress

Table 2 Admission test in relation to fetal distress

Low risk (%)

High risk (%)

Overall(%)

Sensitivity 25 37.5 33.3

SpeciWcity 95.2 94.1 94.7

Positive predictive value 50 75 66.6

Negative predictive value 86.9 76 81.8

123


6–10 at 1 min, 1 continued to be depressed at 5 min also. Inhigh risk group, out of 42 patients with reactive traces, 3(7.1%) had low A/s at 1 min but only 1 (2.3%) wasdepressed at 5 min also. Of the 8 patients with nonreactivetraces, 2 (25%) were depressed at 1 min and continued tobe depressed at 5 min also. Later on both babies expired.Thus the incidence of depressed neonates was higher inmothers with nonreactive traces in both low and high riskgroups (p < 0.05) as shown in Fig. 3. Table 5 shows the sta-tistics.

In the study of 100 cases, in 85 out of 88 (96.59%) withreactive traces the babies were alive and healthy, 13(14.7%) babies required NICU admission and 3 out of 88(3.4%) babies expired due to multiple congenital anoma-lies. In 12 cases with nonreactive traces, 8 (66.6%) babies

were alive and healthy, 4 (33.3%) babies required NICUadmission and 4 (33%) babies expired. In 46 low riskpatients, 3 (6.5%) neonates with reactive traces requirednursery care. In the low risk group with nonreactive traces,1 (25%) required nursery care and 1 (25%) was still borndue to hydrops fetalis with multiple congenital anomalies(renal and cardiac). In the high risk group with reactivetraces, 10 (23.8%) babies required nursery care, 3 babies(7.1%) expired in this group. All 3 babies died due to multi-ple congenital anomalies. Of the 8 high risk patients withnonreactive trace, 5 babies were alive and healthy, 3 neo-nates expired, 2 due to severe birth asphyxia (Fetal Heartrate was 70 at time of admission, so decision of terminationof pregnancy was taken. Even after resuscitation babiescould not be survived.) and 1 died 3 days after birth due tohyaline membrane disease (they did not take up for highercenters). Thus, the incidence of AT to NICU and neonataldeaths was more in patients with non reactive traces in bothrisk groups (p < 0.05) as shown in Tables 6 and 7.

It is observed that out of 100 patients, 27 out of 88(30.6%) with reactive AT and 7 of 12 (58.3%) with nonre-active test had birth weight <2.5 kg. In 46 low risk patientswith reactive traces 10 (21.7%) babies had birth weight<2.5 kg and 1 (25%) baby of mothers with nonreactivetraces had weight <2.5 kg. In high risk group with reactivetraces, 17 (40.4%) patients gave birth to babies of <2.5 kg.6 (75%) of 8 patients with nonreactive traces gave birth tobabies <2.5 kg. Incidentally, it was noted that the birthweight was less in the high risk than the low risk group.

Table 3 Correlation of admission test with mode of delivery (Table 3)

Admission test No. Vaginal LSCS (other than fetal distress)

LSCS (fetal distress)

Forceps Fetal distress

No. % No. % No. % No. % No. %

Low Risk Reactive 46 20 43 16 34.7 9 19.5 1 2.17 10 21.7

Non Reactive 4 1 25 0 0 3 75 0 0 3 75

Total 50 21 42 16 32 12 24 1 0 13 6

High Risk Reactive 42 17 40.4 18 42.8 7 16.6 0 0 7 16.6

Non Reactive 8 4 50 1 12.5 3 37.5 0 0 3 37.5

Total 50 21 42 19 38 10 20 0 0 10 20

Table 4 Admission test in relation to operative interference for fetaldistress

Low risk (%)

High risk(%)

Overall (%)

Sensitivity 23 30 26


Positive predictive value 75 37.5 50

Negative predictive value 78.2 83.3 80.6

Fig. 3 Admission test in relation to APGAR score

Table 5 Correlation of admission test with Apgar score

Low risk (%)

High risk (%)

Overall (%)

Sensitivity 100 66.6 75


Positive predictive value 25 25 25

Negative predictive value 100 97.6 98.8

123

Arch Gynecol Obstet (2010) 282:377–382 381

Occurrence of low birth weight was more common inbabies of patients with nonreactive traces in both the groups(p < 0.05).

Discussion

On analyzing and comparing with other similar studies, it isobserved that in patients with nonreactive traces incidenceof meconium staining was higher, clinically detected fetaldistress was more common, operative interference for fetaldistress was more, incidence of low Apgar score was more,admission to neonatal unit was more and more babies hadlow birth weight in both low risk and hgh risk groups.

In our study, AT was found to have a high speciWcity(94.7%) and a high negative predictive value (NPV)(81.8%) in predicting fetal distress (when a reactive tracewas compared with a nonreactive trace). The sensitivity ofthe test 33.3% and positive predictive value (PPV) was66.6% as shown in Table 8. Ingemarsson et al. [5] in theirstudy found that the incidence of fetal distress was morecommon in the group with an ominous AT when comparedto group with reactive trace (p < 0.0001). They found a sig-niWcantly reduced risk of cesarean section for fetal distresswith a reactive/normal test (RR 0.10, 95% CI 0.03–0.28) in

comparison to those with a suspicious/nonreactive test.They found similar results with a high speciWcity (99.4%)and NPV (98.7%) but low sensitivity (23.5%) and PPV(40.0%). Sarno et al. [6] in their study found a high speci-Wcity (84.5%) and high NPV (98.9%) (similar to our study)in predicting fetal distress. The positive predictive value ofAT in their study was similarly low (23.8%). Their sensitiv-ity was higher (83.3%). In their study found a sensitivity of42.3%, speciWcity of 95.6% and PPV of 73.3% for predict-ing an Apgar score <5 at birth. Sood [7] found in his studythat sensitivity was 41%, speciWcity 94%, positive predic-tive value 83%, negative predictive value 72%, accuracy74% and prevalence 38%. Mires et al. [8] in a RCT, foundno neonatal beneWt, as assessed by metabolic acidosis atdelivery. The odds ratio (OR) for cord arterial blood meta-bolic acidosis was 0.96 (95% CI 0.79–1.17), Apgar score<7 at 5 min 1.07 (95% CI 0.65–1.75), need for IPPV atresuscitation 0.79 (95% CI 0.33–1.85), admission to NICU0.85 (95% CI 0.63–1.14), hypoxic ischemic encephalopa-thy 0.61 (95% CI 0.22–1.65). Women who underwent ATwere more likely to have continuous FHR monitoring inlabor [OR 1.49 (1.26–1.76)] and also more likely to haveoperative delivery for fetal distress [OR 1.36 (1.12–1.65)].This is comparable to the study by Chua et al. [9] whereoperative delivery for fetal distress (p < 0.001), Apgar ·5at 1 min (p < 0.001), Apgar ·7 at 5 min (p < 0.005),assisted ventilation (p < 0.001) and admission to NICU(p < 0.001) were found to be signiWcantly associated withsuspicious or ominous admission CTG. The perinatal out-come was not signiWcantly aVected in either of the twogroups. Chua et al. found thick meconium stained Xuid tobe signiWcantly associated with both increased operativedelivery as well as abnormal fetal outcome.

Table 6 Correlation of admis-sion test with neonatal outcome

Admission test No. Alive and healthy Nursery care Expired

No. % No. % No. %

Low risk Reactive 46 46 100 3 6.5 0 0

Nonreactive 4 3 75 1 25 1 25

Total 50 49 98 4 8 1 2

High risk Reactive 42 39 92.8 10 23.8 3 7.1

Nonreactive 8 5 62.5 3 37.5 3 37.5

Total 50 44 88 13 26 6 12

Table 8 Comparison of the present study with previous studies regarding the ability of admission test in predicting fetal distress

Present study(HCH) (%)

Sarno et al. [6] (%)

Ingemarsson et al. [5] (%)

Sood-Military Hospital, Jodhpur [7] (%)

Sensitivity 33.3 83.3 23.5 41

SpeciWty 94.7 84.5 99.4 94

PPV 66.6 23.8 40.0 86

NPV 81.8 98.9 98.7 72

Table 7 Correlation of admission test with NICU admission

Sensitivity 23.5%

SpeciWcity 90.3%

Positive Predictive Value 33.3%

Negative Predictive Value 85.2%

123


Kulkarni et al. [10] similarly found no signiWcant reduc-tion in risk of reduced Apgar (RR 0.29, 95% CI 0.06–0.74)with “reactive test” as compared to “equivocal” or “omi-nous”. Schifrin et al. [11] also found a higher incidence ofclinical fetal distress with nonreactive traces in both groups.SeraWni et al. [12] reported 12.8% incidence of intrapartumfetal distress in reactive NST and 70% in nonreactive traces.

The limitations of the study are that the present studyincluded only 100 patients in early labor due to resourceand time constraints. A large study may be able to betterevaluate the AT as screening test for predicting adverseperinatal outcome.

In conclusion, AT may be best recommended in allpatients irrespective of whether they are low or high risk asincidence of neonatal morbidity is high in nonreactive trac-ing in centers where advanced facilities are not availableand be cautious. As sensitivity of AT is less in our study, itis not to be recommended as a screening test for everypatient in labor in deciding the mode of delivery as signiW-cant number of patients with nonreactive AT had a normalvaginal delivery fetal outcome. The high speciWcity of theAT means that a normal test accurately excludes adversefetal status at the time of testing. In the absence of acuteevents such as intrapartum asphyxia, an adverse fetal out-come is unlikely if the AT is normal.

ConXict of interest statement None.

References

1. Gabbe SG, Simpson S (2002) Intrapartum Fetal Evaluation,Obstetrics; Normal and Problem Pregnancies. Churchill Living-stone

2. Paul WM, Ouillingan EJ, Maclachlan T (1964) Cardiovascularphenomena associated with fetal head compression. Am J ObstetGynecol 90:824

3. American College of Obstetricians and Gynecologists technicalbulletin (1995) Fetal heart rate patterns: monitoring, interpretationand management No. 207 July 1995. Int J Gynaecol Obstet51(1):65–74

4. FIGO Guidelines for the use of fetal monitoring. Int J ObstetGynecol 1987; 25:159-167

5. Ingemarsson I, Arulkumaran S, Ingemarsson E, Tambyraja RL,Ratnam SS (1986) Admission Test: a screeninng test for fetal dis-tress in labor. Obstet Gynecol 68:800

6. Sarno AP Jr, Ahn MO, Brar HS, Phelan JP, Platt LD (1989) Intra-partum Doppler velocimetry, amniotic Xuid volume and fetal heartrate as predictors of subsequent fetal distress. I. An initial report.Am Obstet Gynecol 161(6 Pt 1):1508–1514

7. Sood AK (2002) Military Hospital Jodhpur, J Obstet GynecolIndia 52(2):71-75

8. Mires G, Williams F, Howie P (2001) RCT of CTG versus Dopp-ler auscultation of fetal heart rate at admission in labor in low riskobstetric population. BMJ 322:1457–1462

9. Chua S, Arulkumaran S, Kurup A, Anandakum C, Selemat N,Ratnam SS (1996) Search for the most predictive tests of fetal wellbeing in early labor. J Perinatol Med 24

10. Kulkarni AA, Shrotri AN (1998) Admission test: a predictive testfor fetal distress in high risk labor. J Obstet Gynaecol Res 24:255

11. Schifrin BS, Lapidus M, Doctor GS, Leviton A (1975) Contractionstress test for antepartum fetal evaluation. Obstet and Gynecol45:433

12. SeraWni P, Lindsay MBJ, Nagey DA, Pupkin MJ, Tseng P,Crenshaw C (1984) Antepartum fetal heart rate response to soundstimulation: The acoustic stimulation test. Am J Obstet Gynecol148:41–45

123

Admission test as precursor of perinatal outcome: a prospective study

Documents

Transcript of Admission test as precursor of perinatal outcome: a prospective study