Effectiveness of adjunct telephone-based postnatal care on maternal and infant illness in the Greater Accra Region, Ghana: a randomized controlled trial

The postnatal period, which begins immediately after childbirth and extends up to 6 weeks (42 days) after birth remains a critical phase in the lives of mothers and their newborn babies [1]. Majority of maternal and infant deaths occur in this vulnerable period. The postnatal period thus provides a critical window of opportunity for delivering lifesaving interventions to both mother and newborn [2]. This is achieved through effective postnatal care (PNC)-care provided for the mothers and newborns from birth of the babies and for the first 6 weeks of life, with the aim of providing optimum health for mother and the newborn babies [1]. Postnatal care is of immense benefit to maternal and newborn health as it creates the opportunity for identification and management of post-natal complications and provide relevant information to mothers on maternal and newborn care during the post-natal period [3, 4].

The World Health Organization (WHO) guidelines on PNC recommends a total of four postnatal visits for every mother and baby as follows: first day (24 hours), third day (48–72 hours), between days 7 and 14, and 6 weeks after birth for health facility deliveries. The first and second postnatal contacts, which are critical for assessing the mother and newborn for complications and teaching the mother essential newborn and selfcare practices, are expected to take place in the health facility prior to discharge, for deliveries occurring in the health facility [5‐7]. Ghana’s PNC schedule consists of at least three contacts for the mother and baby (predischarge care within first 48 hours after delivery, second visit on day 6 or 7 after delivery and last postnatal visit at 6 weeks) [8].

Despite the high PNC coverages in Ghana’s Greater Accra Region, maternal and newborn health outcomes in the region are of great concern, casting doubt on the effectiveness of postnatal interventions. In 2017, 93% of newborns and their mothers received at least one PNC and 87.5% received PNC within the first 48 hours after delivery in the region [9]. In spite of that, neonatal and post-neonatal mortality rates were 19 and 13 per 1000 live births respectively.

Aside being largely client-initiated and heavily dependent on effort of the mothers to seek care [10], usual health facility-based PNC in its current form is over reliant on pre-discharge group counselling and limited to specific scheduled intervals post discharge [1, 11]. These deprive nursing mothers of tailor-made solutions to their problems prior to discharge and uninterrupted access to care while at home. In other instances, particularly in resource-limited settings, mothers are discharged too early. This implies that pre-discharge postnatal contact and counselling, which may be the only postnatal contact received by some mothers, is given too early at the time the mothers are exhausted from childbirth and therefore unable to retain or recall information [11]. Premature discharge of mothers also makes the postpartum length of stay in health facilities too short, leaving insufficient time to recognize and manage complications [12]. Maintaining a continuous linkage with care thus becomes essential for the mothers. In Ghana, however, communication between health workers and mothers throughout the postnatal checks has been observed to be poor [13], with only 23% of nursing mothers surveyed from selected health facilities in the Greater Accra Region reportedly having access to telephone contacts of health workers to clarify concerns during the postnatal period [11].

Though home-based postnatal visit has been used elsewhere to improve maternal and infant health outcomes [3, 14‐17], it is relatively expensive and very difficult to sustain on a large-scale under routine conditions. It does not also guarantee provision of continuously accessible care and information for the mothers and their newborns unlike use of telephone which has produced successful results in some settings [18, 19]. Ghana has a high mobile phone penetration of 131.38% with the Greater Accra Region being the region with the highest mobile phone ownership of 73.7% [20, 21] However, the question that remains unanswered is, whether incorporating telephone follow-up into usual PNC, substantially reduces illness among infants and their mothers in low-resource urban settings. We evaluated the effect of telephone follow-up comprising telephone counselling initiated within 48 hours after discharge and telephone access to a midwife throughout the postnatal period as an adjunct to usual PNC on infant and maternal illness in selected hospitals in the Greater Accra Region of Ghana.

A total of 608 mother-baby pairs were assessed for eligibility, and those eligible were enrolled and studied. Of these, 196 (32.2%) were found to be ineligible and were excluded from the study. Of the remaining 412 eligible participants, 12 declined to participate because they were not interested. The remaining 400 participants were randomized to receive either the usual PNC or telephone-based PNC. Of the 400 participants randomized, 38 (9.5%) were lost to follow-up. In the telephone-based arm, 10 (5.0%) mothers did not receive telephone-based counselling within the planned 48 hours because they could not be reached via telephone. Of the 10 who could not be reached within 48 hours, six (60.0%) were reached after 48 hours and four (40.0%) were never reached and were included in the lost to follow-up. Eight mother-baby pairs from each the study arms did not receive the usual 2 and 6-week PNC because they did not feel the need to. The remaining 362 mother-baby pairs (182 from the telephone-based arm and 180 from the usual care arm) completed the study (Fig. 2). The per-protocol analysis comprised 340 mother-baby pairs: 168 mother-baby pairs in the telephone-based arm and 172 mother-baby pairs in the usual PNC arm whilst the ITT population was based on mother-baby pairs as they were randomized.

This study sought to assess the effectiveness of adjunct telephone-based postnatal PNC on health of infants and their mothers. The study found that compared to their counterparts who received the usual PNC, mother-baby pairs who received adjunct telephone-based PNC were at a much lower risk of falling ill within the first 3 months after delivery. This finding brings to the fore the importance of adoption of mobile health (mHealth) interventions in Ghana’s health care delivery. This is particularly important in the light new wave of telemedicine in Ghana heightened by the ongoing COVID-19 pandemic [26]. In instances in the maternal, newborn and child health continuum of care, where virtual patient care, involving remote exchange of health information between health provider and patient becomes desirable either as a complementary or sole intervention, telephone-based intervention may serve as a foundation.

This study showed that complementing the usual PNC with telephone follow-up significantly reduces the likelihood of illness in both mothers and their babies within 3 months of follow-up. Infants in the intervention arm were significantly less likely than those in the control arm to have fallen ill at least once within the 3-month follow-up period. This indicates that when combined with the usual PNC, telephone-based postnatal follow-up is an effective intervention to reducing infant and maternal morbidity and its attendant consequences of increased expenditure on health as well as overstretched health services. This finding is congruent with findings from a number of other studies conducted in Ecuador and United States of America where significantly fewer infant morbidities were recorded among infants in the telephone-based intervention arm compared to the those in the control arm [18, 19].

This finding is considered an important step towards reducing infant mortality. This is because neonatal mortality remains the main driver of child death globally. Nearly half of all under five deaths (approximately 41%) occur among babies in the neonatal period [27]. Currently in Ghana, whilst infant and under five mortality rates have progressively declined to 21% and 31% respectively from 1990 to 2014, reduction in neonatal deaths has virtually stagnated, with a much slower decline rate of 3% (from 30 to 29 deaths per 1000 live births) [28]. Neonatal mortality thus remains a major contributor to overall mortality of children less than 5 years. Reducing acute illness in the neonatal period is therefore critical in reducing overall child mortality.

While telephone-based PNC significantly reduced the proportion of infant and maternal illness, the effect of the intervention did not vary between subgroups. This finding of consistency of intervention effect across trial subgroups has important implications for practice as decisions on implementation of the intervention will not require considerations of individual characteristics. However, this conclusion has to be given a careful consideration bearing in mind that failure to find heterogeneity across subgroups does not imply definitely that intervention effect seen overall applies to all participants regardless of peculiar characteristics [29]. The lack of subgroup effect seen in this study may be a false negative finding as the study was not powered to determine subgroup effect [30].

The intervention was readily acceptable by the mothers in the intervention arm as evident in their overwhelming expression of satisfaction. Similar findings were observed in Ecuador where almost all the mothers in the telephone intervention arm agreed that the intervention had a positive impact on their lives and they would like to benefit from it in future pregnancies [18]. This finding is however not uncommon as telephone-based interventions in healthcare have received positive feedback from patients and caregivers in many settings [31‐33]. This is probably because of the tremendous flexibility and accessibility it offers. The high level of satisfaction observed in this study, coupled with Ghana’s high mobile phone penetration rate of over 130% [20] and national drive for mHealth [34] favour the intervention as a candidate for adoption and possible scale up as mainstream standard of practice for specific targeted populations.

As a novel strategy of PNC delivery in the Ghana Health Service, adoption and scale-up should be thoroughly thought through taking into consideration first the clinical importance of the effect. Also to be considered in plans for scale-up are availability of human, material and financial resources required to drive the scale-up [35]. Adequate provision of the human and material resources including mobile phones and call credits are key sustainability measures should precede large-scale adoption. Formalizing this intervention has the potential to reduce undue burden on nurses who have been using their personal phones and call credits informally to support patient care [36, 37]. Although a simple intervention which made use of available human resources without additional special training requirements, a cost-effectiveness study is also warranted as a prerequisite for scale-up. Additional studies should also explore variants of the intervention which may either shift the cost of calling to mothers or provide opportunity for cost sharing.

Being an interventional study with a low attrition rate, the study has a high internal validity. However, there are some limitations which must be considered in the interpretation of the results. Firstly, there was a possibility of the Rosenthal effect where outcome assessors might get to know the study arm to which the mothers’ belonged during the interview process and form perceptions which might influence the responses. This was minimized by using different set of research assistants blinded to the allocation intervention for the outcome assessment. Secondly, there was a risk of contamination where participants in the control arm might receive some information from their counterparts in the intervention arm. This was possible because participants interacted among themselves and exchanged telephone contacts while on admission at the hospitals. This risk of contamination in this study was however low as mothers are unlikely to deliver the health education and provide support as trained health workers who are experienced in providing health care. In addition, health workers and their spouses were excluded to minimize co-intervention bias where participants may have received other interventions be it formal or informal which is likely to have contaminated the study. Data on materials used for housing were obtained from interviews rather than observation as the mothers were enrolled at the health facility rather than their homes. Furthermore, both face-to-face and telephone interviews were used to assess outcomes. Though the potential for social desirability bias and measurement errors exists for using this mixed-mode data collection techniques, these are unlikely to affect conclusions drawn from the study as no single mode for outcome assessment was limited to participants in a particular arm of the study. The mixed-mode data collection approach on the other hand, was used to improve response rate and thus reduce the loss to follow-up rate. Finally, being an RCT, the external validity is limited by the urban setting of the trial, the category of health facilities used and the eligibility criteria. Generalizability of this intervention to non-urban settings as well as mothers with characteristics not meeting the eligibility criteria, cannot be guaranteed. Despite these limitations, the study provides potentially useful information that can be reasonably considered for application to targeted populations.

Telephone-based PNC consisting of providing postnatal health education and support beginning 48 hours after discharge plus granting a nursing mother telephone access to a dedicated health worker during the postnatal period when implemented in addition to the usual PNC resulted in statistically significant reduction in proportion of mothers and their babies who fall ill during the first 3 months postpartum. The intervention was largely acceptable to the study participants and shows promise for consideration as an adjunct to usual PNC to improving infant and maternal health outcomes in some settings. Future research is however needed to determine the clinical importance of this finding, its cost effectiveness as well as the optimum format, timing, frequency, and duration of telephone follow-up intervention required to be considered as reasonable option for PNC.