Large cohort study of postnatal events over 18 months in a not-for-profit referral centre in Vellore, South India

Abstract

Objective

To assess maternal medical conditions, physical and surgical ailments, contraceptive use and barriers to its use, maternal mental health, neonatal health, breastfeeding practices and available social support in the postpartum period.

Design

A prospective cohort study.

Setting

A large tertiary care centre.

Participants

12 245 women who delivered after 22 weeks gestation in the year 2022.

Interventions

Three pre-specified exposures, namely mode of delivery, presence of significant risk factors and preterm delivery within the cohort, were used to identify potential groups of women who would need additional support.

The primary and secondary outcome measures

The primary outcome was the number of unscheduled visits by the mother or child and the indications for these visits.

The secondary outcomes in mothers included unhealed wound sites, anaemia, increase in body mass index (BMI) by >3, persistent high blood pressure, pain in the abdomen or pelvis, urinary or bowel problems, musculoskeletal pain, abnormal maternal mental health, breast-related issues and barriers to breastfeeding, contraceptive use and sexual activity.

Results

Only 2% of women and children were lost to follow-up. Nine women and 75 babies died. The majority of infant deaths were related to serious congenital diseases. Unscheduled visits to the health facility were seen in 44% of the cohort, most commonly for upper respiratory infections and fever in the mother and baby. 41 mothers and 741 infants needed admission to hospital. Hospitalisation was more common in those with risk factors or preterm delivery. High blood pressure was seen in 3 to 4% and anaemia in 4% of the cohort. Wound infection was seen in 3 to 4% and urinary incontinence in 2% of women. Wound infection was more common with instrumental delivery. Bowel incontinence was rare. A fourth of the cohort had musculoskeletal pain, especially back pain, which was more common after caesarean delivery. Only 5.5% of the cohort had unsatisfactory mental health, and these women were more likely to have abnormal mental health scores with the NICE Questionnaire at screening. The family APGAR of the cohort was 9/10, and 95% belonged to the middle-income group. 2.6% of neonates had delayed milestones, and this was more common in the group with risk factors and preterm delivery.

Conclusion

Healthcare utilisation was mainly for minor complaints. Re-admissions were rare, as intrapartum and immediate postpartum care were optimal. Women who delivered by caesarean section or delivered a preterm child needed additional support in the postpartum phase. NICE Questionnaire is a quick and easy screening tool to identify unsatisfactory mental health and should be used before discharge, postnatally, even in busy settings. The implementation of formal telephonic support 24 hours a day in birthing facilities should be explored in the future. Holistic postnatal care of mother and child during the immunisation of the baby would be the best opportunity to improve the quality and coverage of care in the postnatal phase.

Trial registration number

CTRI/2022/03/041343.

Strengths and limitations of this study.

• Large single-centre postnatal cohort study that covered all relevant domains of the post-partum phase.

• 12 245 women and offspring followed up for 18 months with very few drop-outs.

• The majority of assessments were done at three time points.

• Only one in-person assessment of mother and child.

• The majority of women belonged to the middle-income group and functional families.

Background

Improving the quality of postnatal care, especially in countries with cultural and economic diversity, has been a challenge. A successful grant to address this important phase in the mother and child enabled the publication of the protocol¹ and the execution of this study. Most current evidence for a positive postnatal experience has been captured in the WHO recommendation² on maternal and newborn care. The postpartum guidelines are not always based on good research³,⁶ and have often evolved with the consensus of experts due to a paucity of data. Information about the postnatal period is mostly from surveys, insurance companies and databases.⁷ Prospective studies with individual patient data are sparse. Discrepancies between discharge preparation and readiness after birth, and the discharge policy have been seen in several countries.⁸ All potential avenues for research in the postnatal period have been highlighted in the research implications⁹ that were described after the development of the WHO recommendations. It has been recognised that as many as 40% of women do not attend a postpartum visit, with attendance rates being lower in populations with limited resources.¹⁰,¹² It is necessary, therefore, to identify all events that warrant the attention of healthcare providers in the postpartum period and the role and timing, if any, of a postpartum visit to a health facility.

The main objectives of our study were to assess maternal medical conditions, physical and surgical ailments, contraceptive use and barriers to its use, maternal mental health, neonatal health, breastfeeding practices and available social support in the postpartum period. These seven domains were chosen for a comprehensive assessment of the post-natal phase towards developing guidelines for pragmatic, cost-effective postpartum care in the year following delivery. The documentation of unscheduled visits by the mother and child, with the indications, enabled us to capture all reasons for anticipated or unanticipated health facility utilisation in this phase. Hospitalisation of the mother and child in the cohort reflected the diagnosis of any serious illness. The three prespecified exposures of caesarean delivery, presence of risk factors and preterm delivery were chosen to identify groups of mothers or babies that would need additional support in the postnatal phase.

Methodology

This large cohort study was executed based on the published protocol.¹ The design, setting, participants, exposure, details of study implementation and strategies for tracing women for follow-up have also been detailed in the protocol.¹

Funding

This study was funded by the Indian Council of Medical Research (ICMR) Department of Health Research, Ministry of Health and Family Welfare, Govt of India. IRIS ID no-2020–9453, RFC-RBMCH/adhoc/63/2020–21. The ICMR had no role in the design, data collection, analysis or interpretation of the study.

Statement of ethics approval

Ethical approval was obtained from Institutional Review Boards (Research and Ethics Committees) of the Christian Medical College, Vellore, Tamil Nadu, India (IRB 12178 dt 24.06.2020), and a detailed written consenting process was followed for each recruited case. The trial has been prospectively registered with CTRI (CTRLink: https://ctri.nic.in/Clinicaltrials/pmaindet2.php?EncHid=NjYyMDE=&Enc=&userName=).

Women who delivered from January to December in the year 2022 after 22 completed weeks in the Department of Obstetrics and Gynaecology, Christian Medical College Hospital, Vellore, Tamil Nadu, were recruited to the study after informed written consent.

All details of baseline data at delivery and in the immediate postpartum period were captured prospectively. They were followed up for 18 months at three contact points, one of which was face-to-face by trained research personnel.

The primary and secondary outcome measures are as follows: The primary outcome was the number of unscheduled visits by the mother or child and the indications for these visits.

The secondary outcomes in mothers included unhealed wound sites, anaemia, increase in BMI, persistent high blood pressure, pain in the abdomen or pelvis, urinary or bowel problems, musculoskeletal pain, abnormal maternal mental health, breast-related issues and barriers to breastfeeding, contraceptive use and sexual activity.

Definitions of some significant variables assessed as outcomes

1. Socio-economic status (SES) was assessed as low, middle and high using the Kuppuswamy socio-economic status scale 2021.¹

2. Family APGAR score of 1 to 10 was used to identify dysfunctional families. Low scores signified dysfunctional families.¹

3. Scheduled visit and special schedule visit (SSV):¹ During the period of our study, all women with no complications were advised to report at around 14 weeks, and this was termed a scheduled postnatal visit. Special scheduling was when the caregiver advised the woman of an earlier visit due to obstetric, medical, surgical, intrapartum or postpartum complications either for the mother or baby.¹

4. Unscheduled visit:¹ Any visit before the scheduled or special visit was termed as an unscheduled visit.

5. Mental health assessment¹:

   1. Risk factor Assessment (RFA) and NICE Questionnaire were used to identify post-natal women with mental illness at baseline assessment.

   2. Edinburgh postnatal depression scale (EPDS) and Perceived stress scale (PSS) were used to diagnose depression.

Care at delivery in this tertiary centre

The labour ward is managed by residents with 2–3 years of experience in the domain of obstetrics and gynaecology. Nurse-midwives only provide additional support but are not primarily responsible for intrapartum care. A consultant with a minimum of 8 years of experience is also present on-site, 24 hours a day. Details of some aspects of the intrapartum care of this setting have been published.¹³ Around 18% of women have an instrumental delivery, with forceps used 80% of the time and vacuum delivery for the rest.

Care in the immediate post-partum period and before discharge

Women are discharged only when they are ambulant, afebrile and have normal bowel and bladder movements. The baby is discharged only when optimal breastfeeding is possible and when there is no loss of weight or hyperbilirubinaemia. Counselling for wound care is given before discharge. The department follows a strict antibiotic policy, with prophylactic use only for Caesarean sections and complications of the third and fourth stages. Antibiotics are started after a blood culture for any patient at the second spike of fever.

Care at discharge

The advice at discharge in our centre is typically informal, and there is no checklist followed. For the duration of this study, we followed some protocols. We advised women to come for a health check only at 12–14 weeks along with the child’s immunisation schedule, as opposed to the usual practice of assessing them at 6 weeks. Baseline details of intrapartum and immediate postpartum events were collated using the postnatal health assessment questionnaire (PNHAQ) using the RedCap software system. This covered important domains of demography, socioeconomic status, family support and social framework. All women had a risk factor assessment (RFA) and the NICE questionnaire completed for mental health assessment as described in the protocol for the duration of the study. This protocol was not practised routinely in the centre. For the study period, we formalised the indication for special scheduled visits (SSV) (Refer to online supplemental appendix), and we also provided each woman with a card listing warning signs for the baby and mother (Refer to online supplemental appendix). This was also not part of the routine care in the department. They were informed that an in-person contact was required at 12–14 weeks for their scheduled postpartum care visit, and there would also be a telephonic assessment at 9 and 18 months.

Patient and public involvement

This study did not involve patients and the public in the design and conduct of the study.

The consenting process, the detailed capture of the demography of families, the mental health assessment and the three follow-up visits were opportunities for unprecedented engagement with the postnatal women and their families. The assessments were done only at the place and times that were convenient for the women. Privacy was ensured at all times during the interviews. The engagement with over 12 000 women at three time points was an opportunity to assess all possible challenges, complications and satisfaction of the women and members of their immediate and extended family in the postnatal care of the woman and her offspring. Thus, this cohort study has been able to capture satisfaction of care through the comprehensive assessment of all outcomes in mother and child, reinforcing public and patient partnership. The outcomes of the study will be disseminated in various appropriate forums after its publication.

Protocol deviation

The protocol was changed to at least one face-to-face contact during any one of the three contacts. This was done as poor compliance was noticed at the face-to-face 14-week visit; we also recognised a difficulty in achieving home visits with the available staff. Details of actual timings of follow-up contact have been captured in the online supplemental appendix. Thus, the timing of actual follow-up visits, especially of the first contact, was delayed and the original time points of 3, 9 and 18 months had to be changed to phases of 6 months.

Data collection

The assessment at these contact points was done by healthcare providers (HCP) /research staff with a nursing or allied health background and with a minimum of 5–10 years of experience in collecting data for research projects. Trained field workers were involved in contacting women and children, checking weight, blood pressure and haemoglobin using a point-of-care test. Good clinical practice certification was ensured for all research staff involved with the study. Interviews at each contact were conducted by a different HCP as a strategy to prevent underreporting of events. Privacy was ensured at all times during the interview. The interviews facilitated the completion of the postnatal health assessment questionnaire (PNHAQ) using the RedCap software system. In one of these visits at the face-to-face assessment, the mother’s weight, blood pressure, haemoglobin and general well-being were checked. Blood pressure was checked using an automated instrument (Omron model HEM 8712-Omron Health Care-India Private Limited-Vietnam). Weight was checked with a Nova weighing machine (Model PS 126-Health Sense Bangalore, Karnataka). These machines were regularly checked for accuracy. Women who attended the parent hospital had haemoglobin checked with an automated spectrometric analyser, and those who had an in-person visit at home had spot haemoglobin estimation using a haemoglobinometer (HaemCue HB 301 Analyser, which was factory calibrated against an international reference). Anaemia was diagnosed when haemoglobin was less than 100 g/L.¹⁴ Postnatal hypertension was diagnosed when the BP exceeded 140/90 mmHg.¹⁵ Maternal mental health assessment was done using the Edinburgh postnatal depression scale (EPDS) and Perceived Stress Scale (PSS) as mentioned in the protocol.¹ Similarly, only at one of the contacts, the baby’s weight and development delay were measured using the Trivandrum Developmental Screening Chart.¹⁶ The details of the baby were also filled in PNHAQ. The women who were lost to follow-up were traced with a minimum of 10 phone calls in order to ensure optimal tracking.

The Statistical Analysis Plan, as described in the protocol was executed. The progression of mothers and infants through the study duration, in terms of number eligible, consented and follow-up, is illustrated using a flow diagram, including reasons for exclusions (Refer to figure 1: Flow chart of women recruited at baseline with follow-up). The frequency of dropouts is presented with reasons when available. Baseline characteristics, as well as demographic data, are presented descriptively.

Figure 1. Flow chart of women recruited at baseline with follow-up. LSCS, lower segment caesarean section.

Descriptive and univariate summary

All variables were analysed as per pre-defined exposure groups, namely mode of delivery (Caesarean vs vaginal), presence or absence of high-risk factors and gestation at delivery (preterm vs term). Within each set of comparisons of pre-defined exposure groups, continuous variables were summarised using mean and SD for normally distributed data, or medians and interquartile ranges for skewed data. Categorical variables were described using absolute frequencies with percentages.

Primary outcome analysis

Frequency of unscheduled events (and other primary outcome variables) was calculated and compared between exposure groups. The χ² or Fisher’s exact tests and unadjusted odds ratios (OR) were used to describe the association between unscheduled events and exposure groups. Continuous outcomes were compared between exposure groups using independent two-sample t-tests or Mann-Whitney tests, as appropriate. Statistical analyses were performed using Stata v11 (StataCorp, Tx, US). Only outcomes that were prespecified as the main comparison of interest (M) and secondary comparison of interest (s) according to the relevance for the exposure pair were analysed, as described in the protocol. Prespecified sub-group analysis was performed for instrumental versus normal delivery, very preterm <32 weeks versus 32–37 weeks and NICE questionnaire assessment >50% at discharge versus <50%. Missing data for contact 1 was less than 0.01 percent; for contact 2, it was less than 0.36 percent and for contact 3, it was less than 0.29 percent.

Results

A total of 12 247 women at 22 weeks or more delivered in the year 2022 in our tertiary centre. Four women died and two women did not consent to participate in the study. 11 267 women had antenatal care in the parent institution and others presented to the centre in labour. The baseline demography and intrapartum details, postnatal complications, maternal and neonatal health, mental health and contraceptive details in relation to the three prespecified exposures have been captured in the online supplemental appendix. Third or fourth stage complications were present in 2039 women in the cohort of 12 245 women. At discharge, 12 241 women were eligible for follow-up. Among 12 577 offsprings of eligible women, 12,271 were eligible to be followed up (241 were stillbirths and 65 early neonatal deaths). Nine maternal deaths, 31 late neonatal deaths (related to serious congenital diagnosis at birth) and 44 infant deaths were identified at follow-up. 41 mothers and 741 infants needed admission during the follow-up. Compliance with SSV was 75%. Details of causes of death, admissions and compliance are captured in the online supplemental appendix. 1989 women were pregnant at the end of the follow-up.

Baseline characteristics were compared according to the three prespecified exposure groups. The risk factors included a spectrum listed in the online supplemental appendix. Most differences among exposure groups, at baseline, could be explained rationally. The the most common indication for SSV in the mother was for suture removal, followed by indications for medical, physical or surgical ailments. Urinary and bowel incontinence was seen in less than 0.5% of the cohort. A permanent method of sterilisation was commonly performed in women undergoing a repeat Caesarean section. Only 3.4% of women used any temporary contraceptive method before delivery. Mental health assessment at baseline identified less than 2.5% of the cohort as high risk for mental illness. Among the neonates eligible for follow-up, 30% were advised SSV, and the most common reason was for prematurity or to monitor the weight gain in the baby. 96% were exclusively breast-fed at discharge. The majority of the cohort belonged to the middle-income group, with 70% of women living in either joint or extended families.

The primary outcome, which was the number of unscheduled visits, did not differ between exposure groups at contact 1 (Refer table 1 and online supplemental appendix). Preterm babies were more likely to have unscheduled visits, with 46% in term deliveries and 50% in preterm deliveries (p=0.004). Fever and wound site infection were the most common causes for unscheduled visits by mothers. Upper respiratory infection, with or without fever, was the most common cause for unscheduled visits for the child. No clinically relevant difference was seen in the primary outcomes at the second and third contacts, for both mother and child (refer online supplemental appendix). Fever and upper respiratory infection were the most common cause for reporting to a health facility, for both mother and child.

Table 1. Primary outcomes – Unscheduled visits to health facility with indications at first contact.

Variables	Total	Mode of delivery			Presence of risk factor		Gestational age
N	12 087	LSCS (n=3969)	Vaginal (n=8118)		Present (n=7730)	Absent (n=4357)	Term (n=10 420) ≥ 37 weeks	Preterm (n=1667)
Unscheduled visit in first contact n (%) (M+B) + (M) + (B)*	5288	1749 (44.07)	3539 (43.61)		3403 (44.04)	1885 (43.26)	4559 (43.76)	729 (43.73)
Number of unscheduled one visit	2913	946 (54.09)	1967 (55.58)		1864 (54.78)	1049 (55.65)	2524 (55.36)	389 (53.36)
2 visits	1234	394 (22.53)	840 (23.74)		783 (23.01)	451 (23.93)	1068 (23.43)	166 (22.77)
3 visits	619	210 (12.01)	409 (11.56)		412 (12.11)	207 (10.98)	519 (11.38)	100 (13.72)
>3 visits	522	199 (11.38)	323 (9.13)		344 (10.11)	178 (9.44)	448 (9.83)	74 (10.15)
For Mother
Mother n (%) (M+B) + (M)*	1735	565 (14.24)	1170 (14.41)		1085 (14.04)	650 (14.92)	1531 (14.69)	204 (12.24)
Fever n (%)1	404	123 (3.10)	281 (3.46)	200 (2.46)	259 (3.35)	145 (3.33)	358 (3.44)	46 (2.76)
				81 (1.00)†
1 Any wound site infection n (%)	332	140 (3.53)	192 (2.37)	94 (1.16)	219 (2.83)	113 (2.59)	305 (2.93)	27 (1.62)
				98 (1.21)†
1 Secondary PPH n (%)	47	18 (0.45)	29 (0.36)	23 (0.29)	33 (0.43)	14 (0.32)	39 (0.37)	8 (0.48)
				6 (0.07†
Redness and tenderness of breast n (%)	87	20 (0.50)	67 (0.83)		46 (0.60)	41 (0.94)	80 (0.77)	7 (0.42)
1 Redness, pain and swelling of one leg n (%)	10	4 (0.10)	6 (0.07)	4 (0.05)	6 (0.08)	4 (0.09)	–	–
				2 (0.02)†
Breathlessness n (%)	8	6 (0.15)	2 (0.02)		6 (0.08)	2 (0.05)	–	–
1 Urinary Problem n (%)	91	38 (0.96)	53 (0.65)	32 (0.39)	–	–	–	–
				21 (0.26)†
Severe headache with visual disturbances n (%)	11	5 (0.13)	6 (0.07)		10 (0.13)	1 (0.02)	8 (0.08)	3 (0.18)
URI/wheezing n (%)	339	119 (3.00)	220 (2.71)		216 (2.79)	123 (2.82)	295 (2.83)	44 (2.64)
Minor Complaints n (%)	602	197 (4.96)	405 (4.99)		371 (4.80)	231 (5.30)	514 (4.93)	88 (5.28)
For Baby
N	12 066	Mode of delivery			Presence of risk factor		Gestational age
		LSCS (n=4119)	Vaginal (n=7947)		Present (n=7771)	Absent (n=4295)	Term (n=10 435)	Preterm (n=1631)
Unscheduled visit n (%) (M+B) +(B)*	5653	1948 (47.29)	3705 (46.62)		3679 (47.34)	1974 (45.96)	4836 (46.34)	817 (50.01)
Baby n (%)	4561	1569 (38.09)	2992 (37.65)		2978 (38.32)	1583 (36.86)	3875 (37.13)	686 (42.06)
2 Fever>100.4/< 97.5 f n (%)	1038	–	–		–	–	868 (8.31)	170 (10.42)	12 (0.73)‡
									158 (9.64)
2 Jaundice beyond 14 days n (%)	27	–	–		–	–	20 (0.18)	7 (0.43)	0 (0.0)‡
									7 (0.43)
2 Abdominal distension n (%)	15	–	–		9 (0.12)	6 (0.14)	12 (0.11)	3 (0.18)	0 (0.0)‡
									3 (0.18)
2 Lethargy n (%)	7	2 (0.05)	5 (0.06)		–	–	5 (0.05)	2 (0.12)	0 (0.0)‡
									2 (0.12)
2 Poor feeding n (%)	32	18 (0.44)	14 (0.18)		–	–	22 (0.21)	10 (0.61)	‡ 2 (0.12)
									8 (0.48)
2 Rapid breathing >60/min n (%)	17	–	–		–	–	14 (0.13)	3 (0.18)	0 (0.0) ‡
									3 (0.18)
2 Noisy breathing, retractions n (%)	30	–	–		–	–	29 (0.28)	1 (0.06)	‡ 0 (0.0)
									1 (0.06)
2 Blue colouring on lips and oral cavity n (%)	3	–	–		–	–	1 (0.01)	2 (0.12)	0 (0.0)‡
									2 (0.12)
2 Foul smelling discharge or bleeding from umbilical cord n (%)	26	6 (0.14)	20 (0.25)		–	–	23 (0.22)	3 (0.18)	0 (0.0)‡
									3 (0.18)
2 Loss of weight n (%)	51	–	–		–	–	39 (0.37)	12 (0.73)	‡ 2 (0.12)
									10 (0.61)
2 Recurrent vomiting blood and bile in vomitus n (%)	27	–	–		–	–	18 (0.17)	9 (0.55)	1 (0.06)‡
									8 (0.48)
2 Crying, irritability and twitching movements without improvement n (%)	91	–	–		–	–	75 (0.75)	16 (0.98)	1 (0.06)‡
									15 (0.92)
Infections n (%)	3368	1165 (28.28)	2203 (27.72)		2221 (28.58)	1147 (26.70)	2882 (27.62)	486 (29.79)
URI n (%)	3106	1078 (26.17)	2028 (25.52)		2049 (26.37)	1057 (24.61)	2646 (25.36)	460 (28.20)
Diarrhoea n (%)	209	81 (1.96)	128 (1.61)		143 (1.84)	66 (1.53)	178 (1.71)	31 (2.00)
Ear infection n (%)	101	28 (0.68)	73 (0.91)		63 (0.81)	38 (0.88)	94 (0.90)	7 (0.43)
Trauma n(%)	–	–	–		–	–	–	–
Others n (%)	128	45 (1.09)	83 (1.04)		92 (1.18)	36 (0.83)	110 (1.05)	18 (1.11)

For subgroup analysis: 1. Instrumental vs normal/vaginal delivery, 2. Very preterm <32 weeks/preterm 32–37 weeks.

Refer online supplemental appendix for test of significance in the table.

^*M+B – Mother and Baby, M - Mother, B-Baby.

^†Instrumental delivery.

^‡Very preterm<32 weeks.

LSCS, lower segment caesarean section; PPH, postpartum haemorrhage; SGA, Sub Group Analysis; URI, upper respiratory infection.

Secondary outcomes in the mother and child, covering the seven domains, are captured in figure 2: secondary outcomes: caesarean delivery versus vaginal delivery, figure 3: secondary outcomes: Presence vs absence of high-risk factor, figure 4: secondary outcomes: term versus preterm delivery (also refer online supplemental appendix for data in tables). The secondary outcomes were prespecified variables, chosen for their relevance to the exposure and the timing of the assessment, to obtain a comprehensive summary of the postnatal phase. There was no difference in anaemia across exposure groups, and the overall prevalence was 4%. Increased blood pressure and increased BMI by 3 were more common in women who had a Caesarean section, or had a risk factor or delivered a preterm baby at any of the three contacts. The prevalence of high BP ranged from 9–11% when checked at the first contact and 3–4% at the last contact.

Figure 2. Secondary Outcomes: Caesarean delivery vs Vaginal delivery. BMI, body mass index; BP, blood pressure; EPDS, Edinburgh postnatal depression scale; PSS, Perceived Stress Scale.

Figure 3. Secondary Outcomes: Presence vs absence of high-risk factor. BMI, body mass index; BP, blood pressure; EPDS, Edinburgh postnatal depression scale; PSS, Perceived Stress Scale.

Figure 4. Secondary Outcomes: Term vs Preterm Delivery. BMI, body mass index; BP, blood pressure; EPDS, Edinburgh postnatal depression scale; PSS, Perceived Stress Scale.

Unhealed or infected wounds were 3–4% in Caesarean wounds and 2–3% in perineal wounds, which were more frequent with instrumental delivery. Abdominal and pelvic pain was most common in the group that had a Caesarean section, but this settled by the third contact assessment. Abnormal vaginal discharge was more common with vaginal delivery, and this again settled by the third contact. Urinary incontinence was seen in 1.4% to 2% of the cohort, and this was not affected by mode of delivery at all three contacts. Women who delivered vaginally, especially those who delivered by instrumental delivery, were more likely to have bowel problems. Constipation was the most common bowel symptom, which settled by the third assessment. Bowel incontinence was extremely rare. Almost a fourth of the cohort had musculoskeletal pain, and a third of these women complained of back pain specifically. It was more common in those who had delivered by Caesarean section compared with those who delivered vaginally. This persisted at all three assessments.

Resumption of sexual activity by second and third contacts was more common in those who delivered vaginally, without risk factors or had delivered a term baby. In subgroup analysis, women who delivered by instrumental delivery were less likely to resume sexual activity in all assessments. 30–40% of women desired contraception. Only 2% of women complained of dyspareunia.

Mental health assessment showed increased stress with abnormal PSS to be more common in the group delivered by lower segment caesarean section (LSCS) only at the first assessment or contact. Abnormal EPDS was seen to be more common in mothers with risk factors and preterm delivery, especially at the first assessment. Unsatisfactory mental health with either abnormal PSS or EPDS was seen in only 5.5% of the cohort, and it was more common if identified by an unsatisfactory NICE questionnaire.

Neonates were more likely to have admissions to a health facility with any significant illness and delayed milestones if there was a risk factor or when delivered prematurely across assessments; 335 neonates (2.6%) were identified to have delayed milestones, and this was more common in the group with risk factors or preterm delivery. The immunisation coverage of infants in the cohort was almost 100%.

Exclusive breastfeeding was seen in 90% of this cohort of women. Women with a Caesarean section, risk factors and preterm delivery were less likely to exclusively breastfeed. Breast-related problems were unusual and mostly seen in the first 3 to 6 months.

Only 20% of the cohort felt they would benefit from a 6-week postnatal visit, and this was more common among women with a Caesarean or preterm delivery. The mean family Apgar was 9 on a score of 0–10 .95% of the cohort belonged to the middle socio-economic group. Some local practices, detrimental to the mother and baby, that have the potential to be incorporated into the health education at discharge were identified (see online supplemental appendix).

Discussion

Our study in the postnatal phase is unique as it has prospectively connected all events in the intrapartum period, with the interphase of the immediate postpartum period to the postpartum phase of 18 months. A comprehensive follow-up of seven domains that cover all aspects of the postnatal period has been achieved. This study has been able to resolve some unknown facts about the postnatal period.

The main findings of this study are that women reported to a health facility for pregnancy-related problems for only about 3 months. Illnesses requiring hospitalisation for both mother and baby were uncommon. Visits to a health facility, both by the mother and the baby, were mainly for fever and upper respiratory infections. Women came to a health facility only if they were significantly symptomatic and not for a general postnatal check-up.

Hypertension, anaemia and wound infection rates were similar to other studies,¹⁷,²¹ confirming that the reporting of health events was optimal. We were very surprised at the very low rates of urinary and bowel incontinence, unlike those seen in previous studies^{22 23} in the literature. These findings may reflect judicious management of labour and the second stage with prudent Caesarean section rates. Contraception was not popular among the women in our cohort and was not considered important, given that the majority of women had only two children. There were fewer terminations of pregnancy, unlike in other studies.^{24 25} . An associated social taboo with termination of pregnancy could be the reason for under-reporting by women, explaining the lower rates in our study. Women with instrumental delivery were less likely to resume sexual activity, but few women complained of dyspareunia, unlike in other studies.²⁶ A significant number of women complaining of back pain postnatally reiterates the need to address this problem in future research. The other surprising finding in our cohort was the low rates of depression in all three assessments, despite using a locally validated instrument, unlike findings in previous studies.^{27 28} However, in subgroup analysis, more women at risk for unsatisfactory mental health using the NICE questionnaire had abnormal mental health scores at assessments. This mental health screening tool was quick and easy to use. It is common knowledge that mental health and social framework and support are intricately related. The majority of the women in our cohort belonged to functional families from a middle socio-economic group, which could explain the low rates of depression in the cohort. Neonatal morbidity and mortality were low in this cohort, and this was reassuring. It was noteworthy that 90% of the cohort of infants were exclusively breastfed even at follow-up. Further research on the undesirable social practices that could harm the mother and baby needs to be pursued locally.

There are some limitations in this study. The assessment by the research HCPs was perceived as the postnatal scheduled visit by some women. The actual weight of the infant was measured, but the adequacy of the weight gain was not calculated. Even though the tally of lost to follow-up was low, some women and children were contacted only once or twice instead of the proposed three assessments. The anonymised dataset of the cohort will be published by Dryad²⁹ to enable public access.

In conclusion, our study is the largest one in the literature that has prospectively captured health events for such a long period. Our study has provided real-time data, unlike other previous studies ³⁰that have indirectly captured some information from electronic medical records and billing data. This cohort of motivated women belonging to the middle-income group can easily represent postnatal women in any part of the world. Our study can be a template for similar studies in different regions of the world to address regional challenges to frame region-specific guidelines. This study has shown that most women and children report to a health facility for minor complaints. Many minor complaints were more common in women who delivered by Caesarean section or had a premature delivery. However, most of these concerns could be easily addressed telephonically. With the use of mobile phones being ubiquitous, it is reasonable to propose formal telephonic support in the postpartum phase instead of an on-site visit. Appropriate research to address the use of telephonic support in the postnatal phase is warranted. NICE Questionnaire is a quick and easy screening tool to identify unsatisfactory mental health and should be used before discharge, postnatally, even in busy settings. Holistic postnatal care of mother and child during the immunisation of the baby would be the best opportunity to improve the quality and coverage of care in the postnatal phase

Thus, the findings of this study can be the platform to address unmet maternal and child health needs in the year following delivery. This could guide the pragmatic timing and methods in outlining the postnatal care programme of the future.

Data availability statement

Data are available in a public, open access repository.