The Cohort of Indonesian Preterm Infants for Long-term Outcomes (CIPTO) study: a protocol

This is a prospective cohort study conducted at the CMGH. All inborn infants are assessed for their eligibility and the participating infants will be followed from birth until discharge (as the baseline data). The sub-categories of preterm birth by the World Health Organization is adopted, i.e.: moderate to late preterm (32 to 37 weeks); very preterm (28 to less than 32 weeks); and extremely preterm (less than 28 weeks). In addition, the gestational age threshold to start treatment of preterm infants is minimum 26 weeks (i.e., a preterm birth less than 26 weeks is currently considered as a nonviable miscarriage). Furthermore, those survival preterm infants will be closely monitored until 8 years old. Outborn infants will be excluded because their data may be incomplete and they might receive inadequate treatments before being admitted into the CMGH [8, 9]. This study had been designed in 2019 and the data collection was started in 2020. As the COVID pandemic hit Indonesia since early 2020, the data collection was temporary withheld. Nevertheless, the data collection was re-started from early 2022. A detailed flowchart of the CIPTO study is presented in Supplementary Fig. 1, while types of post-discharge data collected from the participating subjects and the respective time points are depicted in Fig. 1.

The CMGH is a government-owned public hospital, which also serves as the main academic hospital for the Faculty of Medicine at Universitas Indonesia. The CMGH is a tertiary center serves as the national reference hospital. It has a sophisticated NICU and greater coverage of diagnosis-related group reimbursement from the National Health Insurance, hence current international guidelines could be implemented with several limitations. Maternal and neonatal treatments would be administered to all patients, therefore, according to the recent guidelines without any economic barriers. The CMGH has the most advanced NICU in Indonesia, in which it is supported by a parenteral nutrition center, advanced ventilator devices (e.g., high-frequency oscillatory ventilator and neurally adjusted ventilatory assist), near-infrared spectroscopy, inhaled nitric oxide gas, amplitude-integrated electroencephalography, bedside ultrasound and echocardiography. Furthermore, the management of preterm infants at the CMGH, as an academic hospital, involves other paediatric specialists and other related departments as it is supported by the microbiological and pathological laboratories. The CMGH also able to performed complex surgery involving other departments. This collaboration facilitates routine discussion of cases to ensure that effective treatment shall be provided for each preterm infant.

The inclusion criteria are infants discharged from the neonatal unit of CMGH with (i) born less than 37 weeks of gestational age without any restriction on birth weight; (ii) born at the CMGH; (iii) live in greater area of Jakarta; and (iv) written informed consent obtained from respective parents. The exclusion criteria are (i) multiple congenital anomalies; (ii) suspected of syndrome or chromosomal abnormalities; (iii) parents unwilling to participate; and (iv) self-discharge against medical advice. The drop-out criterion is patients relocate out of Jakarta during the first 24 months of the follow-up period, decline to participate during the study and unable to be contacted during the study period.

Potential predictors in this study are categorized into during hospital admission and after hospital discharge. Potential predictors obtained during hospital admission are gender; gestational age; anthropometric data at birth (weight, length and head circumference); maternal co-morbidities (preeclampsia [10], gestational diabetes [11, 12], intrauterine infection/chorioamnionitis [13], maternal infection and antepartum hemorrhage, as well as COVID-19); fetal distress [14]; history of resuscitation and its related information; antenatal steroid administration [15]; surfactant therapy; type and route of enteral and parenteral feedings; preterm-associated morbidities (sepsis [16], respiratory distress syndrome/hyaline membrane disease [17, 18], necrotizing enterocolitis [19, 20], patent ductus arteriosus [21], intraventricular hemorrhage [22, 23], other intracranial abnormalities, bronchopulmonary dysplasia [24, 25], invasive ventilation use and neonatal seizure [26]); history of blood product transfusion; anthropometric data at discharge (weight, length, head circumference, skinfold thickness and mid-upper arm circumference); parental height; sociodemographic background [27]; family history of prematurity; as well as hearing and vision impairment. Potential predictors after hospital discharge are type and route of feeding; immunization status; any occurrence of infectious disease (both as outpatient and inpatient), as well as growth and developmental assessments during routine visits. A detailed operational definition of each variable is presented in Supplementary Table 1.

The pre-specified long-term outcomes in this study are survival rate; growth outcomes (stunted, wasting, undernutrition, obesity or microcephaly) [28, 29]; feeding behavior [30, 31]; neurodevelopmental outcomes [32, 33]; immunization record according to the current guideline of the Indonesian Paediatric Society [34]; hearing impairment [35]; and vision impairment, particularly retinopathy of prematurity [36]. Survival will be assessed in 6-month intervals by contacting the parents. Growth and neurodevelopmental outcomes will be assessed at 0, 2, 4, 6, 9, 12, 15, 18 and 24 months of corrected age and subsequently at 3, 4, 5, 6, 7 and 8 years old. The growth and development social paediatric specialist team and the nutrition and metabolic disease paediatric specialist team will assess the growth and development at 9, 18 and 24 months old of corrected age and at 3, 5 and 8 years old. The assessment is conducted by trained personnel (with the Developmental Pre-screening Questionnaire as well as Ages and Stages Questionnaire) and verified at the age of 9 months, 18 months and 3 years (with the Bayley Scales of Infant and Toddler Development 3rd edition) and at the age 5 years and 8 years (with the Griffith Scales of Child Development 3rd edition) by a paediatrician or trained psychologist. The Bahasa Indonesia version of those standardized tools are available and have been validated. At least at 9 months of corrected age, each participant will undergo laboratory examination for assessing deficiency of iron or other micronutrient. A referral to other specialist is possible whenever needed (e.g., neurology, respirology, hematology or endocrinology). Feeding behavior will be evaluated by using standardized caregiver-reported questionnaire at 5–6, 9 and 18 months of corrected age. The ear, nose and throat specialists will conduct hearing evaluations, based on the current screening guideline for neonates. A paediatric ophthalmologist will conduct vision evaluation, particularly for retinopathy of prematurity by using RetCam Envision^™ or indirect ophthalmoscope. A detailed operational definition of each variable is presented in Supplementary Table 1. If an impairment is found at any time during the study period, the participant will be referred to receive an optimum care according to the updated local guideline.

Inpatient data will be collected daily and prospectively. Anthropometric measurement is conducted by a trained nurse or physician using standardized tools. Information on parental background, maternal morbidities, preterm-related morbidities and neonatal care will be obtained from the medical records or direct observation. Diagnosis of diseases will be decided by attending physician based on the updated diagnosis criteria. The research team (including undergraduate medical students/paediatric residents/professional officers) will collect the data prior and after the hospital discharge.

After the discharge, patients should have routine monthly health visits during the first year, every 3 months during the second year and every year after 2 years according to guideline by the Indonesia’s Ministry of Health. In this study, we will focus to obtain data at 0, 2, 4, 6, 9, 12, 15, 18 and 24 months of corrected age and annually until 8 years of age. A delayed time of health visit will be rounded to the previous complete months (e.g., home visit at 3 months old is recorded as data at 2 months old). We will communicate with parents routinely to prevent any delay.

The health visit would be performed as scheduled home visit or as routine visit to primary care or paediatric clinic as well as mandatory visit to the CMGH. Data during home visits will be obtained by inspecting the children’s routine health check-up books (direct observation or photographs via mobile phone). As an alternative to home visit, parents could take their children to primary care or paediatric clinic for routine visits and subsequently report the results to the research team. In Jakarta, all primary care clinics have standardized and routinely calibrated anthropometric tools and measurement techniques, under supervision by the provincial and regional health bureaus. The follow-up data on 9, 18 and 24 months of corrected age and 3, 5, and 8 years old will be obtained from the mandatory visits to outpatient clinics at the CMGH. During those visits, the subject will have complete growth and developmental assessments by specialists to validate information obtained during routine home visit.

Data will be stored in internal servers operated by the Center for Health Economics and Policy Studies (CHEPS) at the Faculty of Public Health, Universitas Indonesia. Access for data extraction is only provided to the investigators, in order to protect privacy of all subjects. Any further request for secondary data analysis must obtain permission from all investigators.

Research results will be published in scientific journals and will be presented at scientific conferences as well. Of note, all paediatric clinical guidelines in Indonesia are regularly created and disseminated by the Indonesian Pediatric Society to all pediatricians in Indonesia. There are multiple working units within the Indonesian Pediatric Society, specializing in various health issues of pediatric population. With regard to the preterm infants, the Neonatology working unit is responsible to create relevant guidelines and disseminate them through regular conferences to practicing neonatologists. Results of this study will be shared with the Neonatology working unit to create a reference or guideline to manage surviving preterm infants in Indonesia.

Potential selection bias is noticed due to a decision of not analysing patients who either self-discharged against medical advice or relocated out of Jakarta during the first 24 months of the follow-up period. In order to minimize that bias, close contact with parents will be maintained and the parents will be supported (through parent support group) to prevent self-discharge against medical advice or loss to follow-up. When it is unavoidable, the baseline characteristics between drop-out subjects and study participants will be compared to assess potential confounders or bias sources. The reason of dropping out could be neither related to the observation nor the outcome of interest as well. Thus theoretically would not introduce bias to the data analysis.

Information bias, specifically observation bias, could occur mainly during the anthropometric measurement, although this study will deploy trained personnel and standardized tools. A growth analysis by using percentiles or z-scores might be a way to minimize misclassification bias due to the measurement error. Moreover, paediatric evaluations at 9, 18 and 24 months of corrected age, as well as at 3, 5 and 8 years of age would confirm the previous anthropometric measurements and developmental evaluations conducted in the community setting.

Several minimum sample size calculations were used to provide an adequate statistical power. For estimating the proportion of outcome (incidence) and 10% accuracy within the true value (e), the minimum sample size is 486 subjects, including potential 10% of drop-out and 15% of loss to follow-up. For estimating the survival, the minimum sample size is 193 subjects at the end of 8-year monitoring. For hypothesis testing of predictors with the mentioned study outcomes, the minimum sample size is 251 subjects, including potential 10% of drop-out and 15% of loss to follow-up. All sample size calculations were performed with a formula proposed by Lwanga and Lemeshow [37]. We have decided to recruit at least 500 subjects and the recruitment period requires approximately 2 years (i.e., before 2024).

Descriptive statistics on baseline characteristics of subjects will be visualized in a table or graph. Continuous data with normal distribution will be presented as mean and standard deviation, while the remaining will be presented as median, interquartile range, minimum and maximum. Categorical data will be presented as numbers of observation and percentage. Data distribution analysis using the Shapiro Wilk statistics test with p-value more than 0.05 indicates a normally distributed data. Bivariate analysis of continuous data using independent t-test or one-way ANOVA (parametric test) and Mann-Whitney U or Kolmogorov-Smirnov test (non-parametric test). Categorical data will be analyzed using a 2 × 2 table using the Chi-square test or Fischer’s exact test.

The main effect of interest is the relative risk (RR) and hazard ratio (HR) of pre-specified outcomes. Stratification analysis will be performed to evaluate possible confounder or effect modifier. A confounder is a variable related to predictors, causing the outcomes but not an intermediate variable. Several variables might be treated as predictors or confounders, depending on the analysis of causal relationships. The multivariate analysis will use generalized regression model with an adjusted RR as the main effect or Cox proportional hazard regression with an adjusted HR. Best causal or predictor model of each outcome analysis will be determined based on the most robust statistical model with satisfying discrimination and calibration performance, in which the chosen model will consider the basic disease mechanism. Data transformation or recoding will be conducted if a variable does not meet the criteria for the regression analysis.

Taken together, the CIPTO study is designed to improve long-term outcomes for preterm infants in Indonesia. A close monitoring and rigorous data analyses are expected to produce a new evidence-based reference for preterm care and to improve the overall quality of life of surviving infants in Indonesia.