Indmedica Home | About Indmedica | Medical Jobs | Advertise On Indmedica
Search Indmedica Web
Indmedica - India's premier medical portal

Indian Journal of Community Medicine

Study of Evaluation and Prediction of Neonatal Morbidity and Mortality using ICMR Antenatal Scoring Method

Author(s): Mrudula K.Lala, N.J. Talsania

Vol. 26, No. 4 (2001-10 - 2001-12)

Deptt. of Preventive and Social Medicine, B.J. Medical College, Ahmedabad 380004


Research question: Using ICMR antenatal scoring method, how the occurrence of neonatal morbidity and mortality can be evaluated and predicted.

Objective: To study the evaluation and prediction of neonatal morbidity and mortality using ICMR antenatal scoring method.

Study design: Longitudinal study.

Participants: Pregnant women attending Obstetric O.P.D. at Civil Hospital, Ahmedabad, Gujarat.

Sample size: 1) Registration of 900 pregnant women in first trimester of pregnancy. 2) Follow up of 687 pregnant women till delivery (outcome). 3) 696 neonates i.e. 687 singletons and 9 twins.

Study variables: Age, parity, maternal height and weight, pallor, oedema, blood pressure, Rh status, VDRL results, TT immunisation, factors related to present pregnancy, past pregnancies and past deliveries.

Outcome variables: LBW babies, neonatal morbidity and mortality in relation to ICMR risk score, pattern of occurrence of high risk factors.

Statistical analysis: Chi square test, relative and attributable risk, specificity, sensitivity, positive predictive values of occurrence of adverse outcome.

Results: The incidence of low birth weight babies, neonatal morbidity and mortality increased with increasing risk score. The relative risk of adverse neonatal outcome in pregnancies with risk factors was 5.81 times more than in those without risk factors. The sensitivity of ICMR antenatal scoring system was high (96.33%), while specificity was quite low (20.78%) in predicting the outcome.

Keywords: Neonatal outcome, Neonatal morbidity and mortality, High risk score, Specificity, Sensitivity, Relative risk, Attributable risk


The neonatal period apart from old age is perhaps the most hazardous period in a person's life, not only in developing countries but also in advanced world2. Mortality and morbidity related to this period in our country are still two to three times of those in developed countries3. Occurrence of high risk factors in mother singly or in combination affect the neonatal outcome. There is a system of identifying mothers at risk in our medical practice. However, if this risk can be scored, it will be of immense help in planning and management, both at district and peripheral levels4. ICMR utilized concept of synergism of foetal risk and proposed an antenatal scoring system1,5. In the present study, attempt has been made to correlate neonatal outcome with ICMR antenatal scoring system in hospitalised pregnant women. ICMR scoring is also used to predict neonatal outcome (neonatal morbidity and mortality).

Material and Methods:

Out of 900 women registered in first trimester of pregnancy, 687 (76.33%) women were admitted consecutively and their newborns delivered subsequently in the obstetrics unit of Civil Hospital, Ahmedabad (Gujarat). These 687 women were the subjects of this study. The study period extended from September 1989 to March 1991. Required socio-demographic and obstetric information was collected on predesigned proforma. Few investigations such as haemoglobin estimation, blood grouping, VDRL and complete urine examination were carried out in each case. A numerical value was assigned to each risk factor as proposed by ICMR and the values were added to obtain risk score. The risk score obtained for each pregnant woman at first antenatal visit (First trimester) was updated at second and third trimester and finally on admission to the hospital for delivery. According to scores obtained, pregnant women were grouped as recommended by ICMR1,5. After birth, Apgar scoring and careful physical examination was done for newborns. Birth weight was also recorded.

The outcome was considered adverse if pregnancy ended in neonatal morbidity (defined as any condition requiring admission of newborn to intensive unit) and mortality. Thus neonatal morbidity and mortality of these newborns during hospital stay was noted. All these variables were correlated with ICMR scoring system. Depending upon high risk factors, predictive values for adverse neonatal outcome were assessed using statistical methods4,6.


Table Ia: Risk scoring of certain maternal factors.

Factors Assessment of degree of risk
Maximum (2) Moderate (1) Minimum (0)
Age (Years) 40+ 18 or 30-40 19-30
Parity Prime 5 2-4
Maternal height (cms.) <145 145-150 >150
Maternal weight (Kg.) <40 40-45 >50
Pallor Severe Moderate Nil
Edema Severe & Generalized Moderate Minimum in late pregnancy
Blood Pressure >160/100 >140/90 >120/80
Rh isoimmunisation Rh (Negative) - Rh (Positive)
WR/VDRL Positive - Negative
Tetanus immunization Not done Incomplete Complete.

Figures within parentheses indicate score.

Table Ib: Risk scoring of factors related with pregnancy.

Factors Assessment of degree of risk
Maximum (2) Moderate (1) Minimum (0)
(A) Factors related to present pregnancy:
Breathlessness Yes - No
History of fits in present pregnancy Yes - No
Antepartum haemorrhage Yes - No
Abnormal presentation Yes - No
(B) Factors related to past pregnancies:
No. of abortions >3 2 0
No. of preterm births >2 1 0
No. of still births >1 - 0
No. of neonatal deaths >1 - 0
No. of bouts of bleeding (during or after pregnancy) 2 1 0
(C) Factors related to past deliveries:
No. of past operative deliveries >1 - 0
No. of past difficult deliveries 2 1 0
Hours of abnormal presentation 24 hours 6-24 hours <6

Figures within parentheses indicate score.

ICMR risk scoring revealed that 417(60.70%) women had a risk score 1-3 (mild risk), 139(20.23%) women had risk score 4-6 (moderate risk) and 5(0.73%) had risk score 7+ (severe risk), whereas, remaining 126(18.34%) had a score of '0' considered as "no risk group".

Table II: Risk scoring in relation to mothers delivering low birth weight babies (<2500gm.)

Risk Score Total Births Low birth weight babies (<2500gm.)
(%) Small for date
(%) Total
0 126 3 (2.38) 7 (5.56) 10 (7.94)
1-3 423 47 (11.11) 43 (10.17) 90 (21.28)
4-6 142 20 (14.08) 27 (19.01) 47 (33.10)
7+ 5 1 (20.00) 2 (40.00) 3 (60.00)
Total 696* 71 (10.20) 79 (11.35) 3 (60.00)

*9 pairs of twins; X2=11.47; df=2; p<0.01

On relating ICMR score with different variables of neonatal outcome it was found that incidence of low birth weight (LBW) had direct relation with risk score. It was lowest (7.94%) in women when risk score was '0' and increased with the increase in risk score, being 60% when risk score was 7+. This association was found to be significant. Likewise among LBW babies, occurrence of small for date babies was lowest (5.56%) in '0' risk women and highest (40%) in women with 7+ score.

Table III: Risk score in relation to neonatal morbidity and mortality.

Risk Score Total Births Neonatal
(%) Neonatal
(per 1000 births).
0 126 3 (2.38) 1 (7.94)
1-3 423 45 (10.64) 21 (49.65)
4-6 142 28 (19.72) 9 (63.38)
7+ 5 2 (40.00) 0 (0)
Total 696 78 (11.21) 31 (44.54)

X2=24.83; df=2; p<0.001

There were 696 live births, out of these, 78(11.21%) developed significant illness during hospital stay and of them 31(4.45%) died within a week after birth. This neonatal morbidity and mortality showed direct relation with risk score. Neonatal morbidity was 2.38% in '0' risk women; it gradually increased, as risk score increased and was highest (40%) in severe risk women. This relationship between neonatal morbidity and ICMR risk score was found statistically significant. Neonatal mortality rate in '0' risk group was lowest 7.94 per thousand births and it was highest (63.38) in 4-6 risk group (Table III).

Table IV: Occurrence of high risk factors and adverse outcome.

Risk Factor No. of Women % Neonatal morbidity p value
Absent 468 0.64 <0.05
Present 219 2.74  
Maternal Weight (Kg.)
>45 496 0.60 <0.001
<45 191 7.33 <0.001
H/o 2 or more successive abortions
No 594 0.51 <0.001
Yes 93 9.68 <0.001
Pregnancy order 1
Yes 517 0.58 0.05
No 170 2.94 0.05
Pregnancy order>5
No 632 0.47 <0.001
Yes 55 14.55 <0.001
Maternal Age (Years)
<30 595 0.50 <0.001
>30 92 9.78 <0.001
Maternal Age (Years)
>18 667 0.45 <0.0001
<18 20 35.00 <0.0001
H/o 1 preterm birth
No 604 0.50 <0.0001
Yes 92 13.42 <0.0001
H/o Neonatal death(s) in previous pregnancy*
No 447 0.55 <0.001
Yes 70 7.14 <0.001
H/o still birth(s) in previous pregnancy*
No 491 2.20 <0.001
Yes 26 16.13 <0.001

* 170 women of first pregnancy order were excluded.

Table IV lists ten risk factors and their level of significance with neonatal morbidity. All the ten factors were associated with occurrence of neonatal morbidity. It is evident from Table IV that there was increased risk of neonatal morbidity when maternal weight was below 45 kg, high pregnancy order, high maternal age at delivery and previous history of stillbirth (p<0.001). The relation of low maternal age at delivery and history of one previous preterm birth with neonatal morbidity was highly significant (p<0.0001) while relation of presence of pallor and first pregnancy order with neonatal morbidity was just significant (p<0.05).

Table V: Relative and attributable risks associated with or without maternal risk factors for occurrence of neonatal morbidity and mortality.

Maternal Risk factors Total Maternity Adverse Outcome Occurrence of adverse outcome % Relative risk Attributable risk
Neonatal Morbidity
Present 570 75 13.16 5.54 81.96
Absent 126 3 2.38    
Neonatal Mortality
Present 570 30 5.26 6.66 84.98
Absent 126 1 0.79    
Total Outcome
Present 570 105 18.42 5.81 82.79
Absent 126 4 3.17    

Table V illustrates that relative risk of neonatal morbidity and mortality was 5.54 and 6.66 times more respectively in pregnant women with risk factors compared with women without risk factors. This is highly significant (p<0.001). Table also reveals that attributable risk of neonatal morbidity and mortality associated with maternal risk factors was 81.96% and 84.98% respectively while relative risk and attributable risk for total neonatal outcome was 5.81 and 82.79%.

Table VI: Analysis of scoring system for prediction of adverse neonatal outcome.

Adverse outcome Sensitivity (%) Specificity (%) Positive predictive value (%)
Neonatal morbidity 94.87 20.16 13.19
Neonatal Mortality 96.77 18.80 5.26
Total Neonatal adverse outcome 96.33 20.78 18.42

Table VI describes sensitivity, specificity and positive predictive value for the total pregnant women for neonatal morbidity and mortality. The sensitivity (outcome predicted as high risk) was high (96.33%), whereas, specificity (no neonatal morbidity and mortality at low risk) was low (20.78%) and positive predictive value (Proportion of women with risk factors later turn out to have neonatal morbidity and mortality) was quite low (18.42%) for neonatal out come. The sensitivity, specificity and positive predictive values were 94.87%, 20.16% and 13.19% respectively for neonatal morbidity. While sensitivity, specificity and positive predictive value for neonatal mortality were 96.77%, 18.80% and 5.26% respectively.


Correlating ICMR risk score with different variables of neonatal outcome; direct relation was observed. The incidence of LBW increased as risk score increased. Similar were the observations of Rege et al7 and V.Krishnan et al5. V.Krishnan5 had used same ICMR scoring system in his study. In our study8, it was also observed that as risk increased, postnatal complications and occurrence of complicated deliveries increased8.

Neonatal mortality rate was 44.54 per thousand births. V. Krishnan5 reported neonatal mortality rate as 47.7 per thousand live births while Singhal et al9 found an neonatal mortality rate of 21.8 per thousand live births in his study. The neonatal mortality rate was maximum in women with severe risk and minimum in women with no risk. A similar pattern of neonatal deaths with ICMR score was reported by V.Krishnan5 in his study at Dufferin Hospital, Lucknow. Neonatal morbidity rate was 11.21% in the present study, well compared with other studies10.

Analysis of individual risk factors in relation to adverse neonatal outcome showed maternal malnutrition, higher pregnancy order, extremes of age at maternity, previous history of preterm deliveries and foetal loss as contributing significantly in predicting neonatal morbidity in newborns (Table IV). Pallor, first pregnancy order and history of neonatal death(s) in previous pregnancy also contributed to predicting neonatal morbidity. In the community based study of Reddaiah V.P.11 relative risk to adverse outcome was 4.69 times more than those without risk factor, while sensitivity and predictability were very low as 29.6% and 17.8% respectively and specificity was very high (92.9%) for perinatal loss11. In the study of Ebernard M.H.12 the specificity was (84.9%) quite high and positive predictive value (21.9%) and sensitivity (38.5%) were quite low for occurrence of preterm births.

Our findings did not compare with above studies11,12. The relative risk of neonatal adverse outcome was 5.81 times more in high risks group. The sensitivity (96.33%) was quite high, while the positive predictive value (18.42%) and the specificity were quite low (20.78). In our published study13, the specificity and positive predictive value were low and sensitivity was high for preterm deaths and perinatal loss. Same pattern of findings were observed for neonatal outcome in our study.

Thus, ICMR scoring system is good in predicting neonatal morbidity and mortality by identifying high-risk conditions (i.e. its sensitivity is high). But on the other side, this system is poor at predicting healthy newborn in low risk women because specificity is very low (20.78%). This difference might be due to use of different scoring systems in different studies. The other reason may be, our study was a hospital-based study and most of the pregnant women availing hospital facilities belonged to low socio-economic group. This might be responsible for very low percentage of pregnant women (18.34%) scored as no risk. Moreover, pregnant women belonging to low socio-economic group did not attend antenatal clinics regularly and when complications occurred they were hospitalized. In such hospitalized cases, neonatal outcome is always bad (sensitivity high). As the specificity was low (20.78%) in our study it was concluded that for pregnant women at no risk attending hospitals, regular antenatal check-up should be advised to detect complications and to avoid adverse neonatal outcome.

Conclusions and Recommendations:

It can be concluded from this study that by introducing simple system of risk scoring during antenatal check up, all pregnant mothers can be placed in the different slots of mild, moderate or severe risk. Accordingly, further care and referral services can be imparted to them and care of mother and baby can be started immediately. ICMR risk score technique can be used at all levels of health care. This will help us in lowering neonatal mortality and morbidity rates and achieving our goal of Health for All. In hospitals catering low socio-economic population; high risk pregnant women (81.66%) are more, but attention should also be given to no risk or less risk cases. There is need to carry out community based studies by applying ICMR risk score.


The authors are thankful for financial assistance rendered by Additional Director Medical Education and Research, Gandhinagar, Gujarat State.


  1. ICMR comprehensive care of mother and children (Intervention Phase) written communication,; ICMR, New Delhi; 1986.
  2. Sheila A. Bhave. Trends in perinatal and neonatal mortality and morbidity in India. Indian Paediatrics 1989; 26, 11: 1094.
  3. Rekha Harish, Misra PK, Malik GK. Goodwin's high risk score and neonatal outcome: An evaluation; Indian Paediatrics, 1989; 26: 800.
  4. World Health Organisation. The risk approach in health care, Public Health Paper, 1984; No. 76.
  5. Krishnan V, Idris M Z, Srivastava VK et al. Scoring of high risk mothers and related outcome. Indian Journal of community Medicine 1988; XIII(4): 176-9.
  6. Austir Bradford Hill & I.D. HIM Bradford Hill S. Principles of Medical Statistics, Twelfth Edition, 1993: 132-4.
  7. Rege JS, Mehba AC, Patel DN. Simple clinical monitoring in pregnancy Part II, Journal of Obstetrics & Gynaecology India, 1985; 35: 476-80.
  8. Niti J.Talsania, MK Lala. Scoring of high risk pregnant women and related outcome. Indian Journal of Maternal and Child Health, 1990; 2(3): 92-4.
  9. Singhal PK, Mathur GP, Mathur S, Singh YD. Perinatal mortality in ICDS urban slum area, Indian Paediatrics, 1986; 23: 343.
  10. Pandit Anand, Bhave Sheila. Developmental assessment, follow up and intervention in high risk neonates, Report of proceedings of workshop held at K.E.M. Hospital Pune, under National Neonatology Forum, 1990: 1-44.
  11. Reddaiah VP, Kappor SK. Risk approaching maternal care. How beneficial is this approach in reality? Indian Journal of Paediatrics 1985; 52: 61-5.
  12. Eberhard MH, David SG. Evaluation of risk scoring in preterm birth prevention study of indigent patients American Journal Obstetrics & Gynaecology, 1989; 4: 829-32.
  13. NJ Talsania, MK Lala. Evaluation of antenatal risk scoring in a preterm birth prevention and perinatal loss. Indian Journal of MCH 1994; 5(1): 5-9.
Access free medical resources from Wiley-Blackwell now!

About Indmedica - Conditions of Usage - Advertise On Indmedica - Contact Us

Copyright © 2005 Indmedica