Ethical approvals

The study was approved by the institutional ethics committees of each participating institution and the WHO Ethics Review Committee (Protocol ID NCH09008). Written informed consent was obtained from caregivers for each activity.

Interventions/Major activities and sub-activities

The interventions were classified into four main categories.

1. Intervention 1.0: Home-based care included 3 sub-activities
   • Sub-activity 1.1: Community Health Workers (CHWs)/ Community Health Extension Workers (CHEWs) carried out community-based surveillance to identify pregnant women. The visits were calculated based on the total number of pregnant women identified in one year as compared to the whole period and total surveillance visits reported by the sites.
   • Sub-activity 1.2: CHWs/CHEWs made prenatal visits for health promotion, exclusive breastfeeding, seeking care from skilled birth attendants during pregnancy and delivery, preparation for delivery, prevention of malaria and promoting good dietary habits. During a prenatal visit, 10% of the provider time was assumed for educating the mother about the danger signs.
   • Sub-activity 1.3: Ten postnatal home visits were to be made on days 1, 3, 7, 14, 21, 28, 35, 42, 49 and 60 to promote optimal care practices such as breastfeeding, keeping the baby warm and hygiene; to identify danger signs in mothers and newborns, and to promote appropriate care-seeking.
2. Intervention 2.0: Link between CHWs and nurses was classified under two sub-activities
   • Sub-activity 2.1: Once a CHW/CHEW identified a young infant with any danger sign, the young infant was referred either to a hospital or to the health center nurse. The sick young infants were taken to a health center by CHWs or visited by the nurse at home.
   • Sub-activity 2.2: CHWs/CHEWs revisited the homes to check the outcome of the referral or treatment.
3. Intervention 3.0: Assessment and management of sick young infants using Integrated Management of Childhood Illness (IMCI): The nurse assessed the sick young infants, either brought directly by the mother/caregiver or through the CHW/CHEW. All young infants who had any danger signs were assessed for possible serious bacterial infection (PSBI) (sub-activity 3.1). Those with PSBI signs who accepted referral to a hospital for further management, were counseled and prepared for referral (sub-activity 3.2).
4. Intervention 4.0: Outpatient treatment—If the family refused referral, young infants with PSBI signs were reclassified into three categories: 1. Critically ill young infants with signs such as unconsciousness, convulsions, unable to feed at all, apnea, unable to cry, cyanosis, bulging fontanel, persistent vomiting (defined as vomiting following three attempts to feed the baby within 30 minutes) and weight < 1500 g at the time of presentation were again referred to a hospital and not enrolled in the study. 2. All young infants classified as having CSI were enrolled in the trial after consent was obtained, and randomized to either the reference therapy (regimen A) or one of the experimental treatment regimens (B, C or D) (Box 1) for outpatient treatment (sub-activities 4.2 to 4.5). 3. Young infants with only fast breathing whose families refused referral were classified as having pneumonia, and after obtaining consent, were randomized to either regimen A or oral amoxicillin (regimen E) (Box 1) for outpatient treatment (sub-activity 4.6). All enrolled young infants were assessed daily (sub-activity 4.1). In the research setting, independent outcome assessors, who were experienced nurses, visited the homes of enrolled young infants on days 4, 8, 11 and 15. However, in the government setting only one visit is expected by the CHW/CHEW, and this cost was included under sub-activity 2.2. If the treatment failed, the young infant was referred to a hospital. If the referral was refused by the family, rescue treatment with injectable ceftriaxone for seven days was given by independent outcome assessors on an outpatient basis. The cost of rescue treatment has been excluded from this study.

Outpatient services by a nurse for sick young infants, such as administering injectables or assessment of non-response to treatment, was provided at government clinics in the DRC and Kenya. In about 10–15% of cases, a nurse made a home visit and administered the indicated injections. CHWs supervised administration of the first dose of oral amoxicillin every day at the home of the infant, while the second dose was given by the parent. In Nigeria, CHEWs initially identified sick young infants in the community and referred them to the nurse at a health center for assessment, enrollment, randomization and provision of the first injectable dose of treatment. Thereafter, the CHEW administered the first dose of oral amoxicillin daily and provided injectable therapy at the home of the young infant. The second dose of oral amoxicillin was given by the parent.

Estimation of covered and treated young infants

The number of young infants covered was estimated by adding the total number of young infants that had at least one postnatal visit. Treated young infants were estimated by adding those who had at least one day of treatment under any regimen. The total number of visits for 1.3 and 4.1–4.7 were calculated by multiplying the covered or treated young infants by the number of days of visits per infant. For example, under regimen B in Kenya, if one young infant was treated for two days, one young infant for three days, seven young infants for four days, and 119 young infants for seven days, then the total number of young infants treated would be 1+1+7+119 = 128 and visits would be 866 = (2*1)+(3*1)+(4*7)+(7*119). Similarly, the number of visits for other activities was determined by the number of women or young infants receiving the intervention and the number of days of visits for that activity.

Estimation of direct and indirect costs

Direct costs for the activities and outpatient treatment under each regimen described above were calculated per young infant treated by adding the per-visit cost for human resources (opportunity costs of providers); drugs and consumables costs for complete seven-day treatment and an incomplete three-day treatment (averaged for 1–6 days of treatment). Incomplete/failed treatment happened either because the young infant did not respond, was withdrawn from treatment, or died on any of the 1–6 days of treatment. The direct total costs for outpatient treatment under each regimen were calculated by multiplying per young infant cost with the number of young infants receiving 7 days or 3 days treatment.

The direct provider costs under any regimen included the pre-outpatient (Pre-OP) costs and outpatient (OP) treatment costs for that regimen. Direct human resource costs for each activity and regimen were determined based on the total duration spent for a visit and an episode (estimated based on times per day and days per young infant) of treatment and salary of the providers (per minute gross income estimated based on 25 working days in a month for 8 hours a day). The costs of home-based care (Intervention 1.0), links to a facility (Intervention 2.0), assessment and management (Intervention 3.0) and daily assessment of enrolled young infants (sub-activity 4.1) were calculated only based on human resource costs per visit and the total number of visits. Based on expert opinion, only 10% of the human resource costs for sub-activities 1.1 and 1.2, and 20% of the costs for sub-activity 1.3 were included as direct provider costs for PSBI treatment and management.

Costs of medicines and consumables were calculated per administration or per treatment course by multiplying the quantities required with prices. While the injections costs were estimated per administration, oral amoxicillin was calculated per treatment course. As unused amounts after dilution were not used later, the quantities for treatment with oral amoxicillin remained the same with 14, 10 or 4 doses under different regimens. Based on the number of administrations per day and number of days of treatment for each regimen, total amounts of drugs and consumables required for a full 7-day and 3-day treatment under different regimens were calculated. The price of drugs and consumables was also calculated per dose for full seven days or partial three days of treatment.

The costs of equipment and capital were estimated by using both the depreciated and discounted rate. A social discount rate of 6% was used for discounting equipment and vehicles [26]. Marginal financial costs were estimated for indirect cost items such as non-consumables, operations, training and personnel costs for supervision and administration necessary for running the program effectively. Items included under direct and indirect cost categories for different interventions, and programmatic and administrative activities are shown in Table 1. Items considered as part of the research and not routine scaleup, already existing part of the health system or one-off start-up/ introductory costs were not included as part of the cost calculations.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Items included under direct costs, indirect costs and research/ health systems/ start-up costs for each activity.

https://doi.org/10.1371/journal.pone.0247977.t001

The Pre-OP direct costs per treated young infant under interventions 1, 2,and 3 were estimated by adding the total direct costs under these interventions and then dividing by the total number of young infants treated under any regimen at a given site. Direct OP costs per treated young infants for each regimen were estimated by adding the costs for daily assessment under activity 4.1 to each regimen and then dividing the cost for that regimen with the number of young infants treated under that regimen at a given site. Direct costs per covered young infant for any intervention or regimen was estimated by dividing that intervention costs by the total number of covered young infants at that site. Indirect costs per treated or covered young infant were estimated by dividing the total costs with the number of infants covered at that site. Finally, the direct (Pre-OP and OP) and indirect costs were added to estimate the total treatment costs under any regimen per treated young infant and covered young infant.

Average costs for scaleup for each regimen were obtained by taking the weighted average of costs across different sites. Weights were the numbers of treated young infants in 2012 at each site. For average costs, 95% confidence interval (CI) were calculated.

Estimation of incremental cost-effectiveness ratios

The effectiveness indicator was calculated using the risk difference in treatment failure and was estimated as the percentage of newborns who did not fail treatment after one week of enrollment with each of the regimens. Treatment failure for fast breathing was defined as death; clinical deterioration; hospitalization; persistence of fast breathing on day 4 or recurrence after day 4 up to day 8 of enrolment; development of a serious adverse event other than death that is related to the study antibiotics [14]. Whereas, the treatment failure definition for the CSI was death; clinical deterioration; hospitalization; no improvement in clinical condition by day 4; persistence of any sign of CSI on day 8 of enrolment; development of a serious adverse event other than death that is thought to be related to the study antibiotics [15]. Using the percent risk difference in treatment outcomes reported in [14, 15] for the whole period and all enrolled young infants, effectiveness percentage for each regimen was calculated and then incremental effectiveness was calculated against regimen A. Using weighted average cost per treated young infant across the sites, the incremental costs were calculated for each regimen by taking the difference in the costs of the given regimen against regimen A. The ratio of incremental costs to incremental effectiveness provided the incremental cost-effectiveness ratios (ICER). Comparisons were made for experimental regimens B, C and D against the regimen A for CSI and experimental regimen E against the regimen A for fast breathing. Pairwise comparisons were also made between regimens B, C and D. If the ICER was less than 0 and the value lies in the right lower quadrant (that is increased effectiveness and lower costs), then the option was considered cost-effective as compared to the reference regimen.

Data collection

All the data were directly collected by the researchers implementing the study on the site (primary data from facilities and providers). Data were collected in pre-prepared forms from each site on the number of visits for each sub-activity and number of young infants covered and treated under each intervention. For this study, the data on the total number of young infants for each intervention and number of providers were directly obtained from the main database for 2012. The costing data for 2012 included 39% (2310 out of 5897) of young infants reported as treated with different regimens during the whole period of the study (from April 2011-June 2013) (S3 Table).

For human resources, a time and motion study was carried out to collect data on duration (in minutes) for each sub-activity for each type of provider, as self-recorded on questionnaires specifically prepared for this purpose [27]. Approximately 25% of each type of providers were also followed by an interviewer on a random day to validate the self-recorded data. The number of CHW/CHEWS and nurses at each site and those surveyed are shown in S4 Table. Overall, 83% of the providers were surveyed, with 71% of all nurses and 87% of CHW/CHEWs surveyed.

Time spent on travel, waiting and visits where no contact could be made with the caregiver was recorded separately and split across the sub-activities depending on the purpose of the travel. Personal time was not included. The average time taken by CHWs, CHEWs and nurses was estimated per visit per woman or young infant for which the sub-activity was undertaken. If during a visit of 20 minutes, four women were surveyed in a household, then five minutes were estimated per visit per woman. If more than one provider was involved in an activity, then the weighted average of provider cost was calculated. For example, in Nigeria, the first dose of oral amoxicillin was given by the nurse at the clinic and the subsequent six first daily doses by a CHEW at home. For a 7-day treatment with two doses per day, 1/7 (14%) was provided by the nurse and 86% was by the CHEW. The second dose was given by the parent each day, and no human resource costs were attributed. Data on salaries prevalent at the sites were collected from official records and average salaries were taken including the emoluments.

Data on quantities of drugs and consumables were collected based on per administration or per episode of treatment for each regimen. Based on the number of administrations per day and number of days of treatment, total amounts of drugs and consumables required for a full 7-day treatment under different regimens are shown in S2 Table. Drugs were centrally provided through WHO, and international procurement prices were used when local prices were not available. In the DRC and Kenya local generic prices were used while international procurement prices were used for Nigeria. Local prices for all consumables were collected from local markets. The prices of drugs and consumables used across different sites are given in S5 Table.

For indirect costs, data on the percentage of time spent each day and number of days in a month were collected from all personnel involved in administrative and supervisory activities (managers, supervisors, CHW/CHEWS and nurses) in the covered area. As several of these were research and health system staff, marginal costs calculations included only the staff that was considered essential for conducting the government program as shown in Table 1. Similarly, for capital equipment and other non-consumables, even though the data was collected for all consumables and capital equipment purchased, only the items that were an important part of the program costs were included (Table 1).

Excel was used for data analysis. The exchange rates used for converting local currency units to US$ for 2012 are 910 Francs (DRC); 86.1 Shillings (Kenya); and 156.15 Naira (Nigeria) [https://www.xe.com/ accessed 2.12.2018]. The average cost estimates are also discussed in 2018 prices using the latest available Gross Domestic Prices (GDP) deflators from International Monetary Fund database [https://data.imf.org/regular.aspx?key=61545852 accessed 15.11.20].

Covered and treated population; staff time for each activity

Table 2 provides the number of visits and treated young infants for each activity, and covered young infants for each site. The highest number of assessed and covered young infants were in Ile-Ife at 12,441 and the lowest in Zaria at 4406. In terms of those enrolled and treated, the highest numbers were in Kenya at 646 as compared to 343 in Zaria. The percentage of infants treated out of those covered was the highest in DRC (8.8%), followed by Zaria (7.8%); Kenya (6.6%); Ibadan (4.7%) and Ile Ife (3.3%).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Number of visits for each intervention and activity; the number of treated young infants under different regimens and the number of covered young infants at five sites in DRC, Kenya and Nigeria, 2012.

https://doi.org/10.1371/journal.pone.0247977.t002

To calculate the human resource costs, data on duration (in minutes), place of activity i.e., in the health center or as an outreach activity and percentage of total visits covered by a specific provider for each activity (in brackets) are shown in Table 3. While the activities under interventions 1, 2, 3 and 4.1 in Table 3 correspond to the activities in Table 2, the treatment under different regimens (activities 4.2–4.6) required a combination of sub-activities (4a-4c) shown in Table 3. Average monthly salaries of CHW/CHEWS and nurses at different sites are also shown in Table 3. It can be seen that the average time taken by providers for several outreach activities was much higher in DRC, due to longer distances between communities from the facilities and mostly the providers travelled by foot or bicycles.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Average time taken in minutes for each provider per visit by activity and monthly salaries of the providers at five sites in the DRC, Kenya and Nigeria, 2012.

https://doi.org/10.1371/journal.pone.0247977.t003

Cost estimation for regimens

Disaggregated direct economic costs per young infant treated for a complete seven days of treatment under each regimen are shown in Table 4. The human resource costs per young infant for daily assessment were added under each regimen. The costs of seven days of treatment per young infant treated showed that the costs of human resources were higher for regimen A in the DRC and Kenya and regimen B in Nigeria. The drug and consumables costs were higher for regimen A, followed by regimen B, C, D and E respectively at any given site as injectables were more expensive compared to the oral treatment. The lowest human resource costs in the DRC were primarily due to the lower salaries and also because most injectables were provided in the health centers so that the travel and waiting time for nurses were reduced. In Kenya, the regimen that involved injectables administered by a nurse was more expensive due to higher salaries. In Nigeria, CHEWs replaced many functions performed by nurses, such as the continuation of injectable therapy (after the first injection), so the cost of treatment was lower due to lower salaries of CHEWs than registered nurses. However, this lower cost was offset as all treatment except the diagnosis and the first antibiotic dose in Nigeria was an outreach activity by CHEWs, requiring more time than treatment at a health center. Drug costs across sites in Nigeria were similar as international procurement prices were used for medicines but were lower in the DRC and Kenya where price for generic medicines were used (S5 Table). Kenya had higher costs for consumables (needles and distilled water vials) as compared to all other sites.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Costs of human resources, drugs and consumables per young infant treated for complete seven days treatment under each regimen at five sites in the DRC, Kenya and Nigeria, 2012, US$.

https://doi.org/10.1371/journal.pone.0247977.t004

The total cost of treatment including pre-outpatient activities; outpatient treatment costs under different regimen (for 7 days or less); and indirect costs for administration and program operations per young infant treated and per young infant covered across different treatment regimens for each site are shown in Table 5. The cost per young infant with CSI treated with different regimens were highest for reference regimen A (US$ 16 in the DRC to US$ 31 in Ile Ife and Ibadan in Nigeria) and lowest for regimen D (US$ 11 in the DRC to US$ 25 in Nigeria). Treatment for fast breathing alone (regimen E) varied from a low of US$ 10 in the DRC to high of US$ 24 in Ile Ife, Nigeria. The weighted average costs for CSI in 2012, when averaged across all sites, were US$ 25.3 (95% CI US$ 20.4–30.1) with regimen B, US$ 22.4 (95% CI US$ 18.0–26.7) with regimen C and US$ 20.9 (95% CI US$16.4–25.3) with regimen D compared to reference regimen A at US$ 27.4 (95% CI US$ 22.5–32.3). The weighted average cost for regimen E for fast breathing was US$ 18.3 (95% CI US$ 13.4–23.3). The corresponding costs per treated young infant in 2018, when adjusted for inflation, were US$ 43.2 for regimen A, US$ 39.3 for regimen B, US$ 34.9 for regimen C, US$ 32.5 for regimen D, and US$ 29.0 for regimen E.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. Pre-outpatient (Pre-OP), outpatient (OP), indirect/administrative and total costs per treated and covered young infant^* for five regimens for PSBI in the Democratic Republic of Congo (DRC), Kenya and Nigeria, 2012, US$.

https://doi.org/10.1371/journal.pone.0247977.t005

The indirect administrative costs per treated young infant varied between US$ 4.9- US$ 7.5 across sites. Indirect programmatic and administrative costs, estimated based on maximum treatment costs, ranged between 22% to 46% (highest being in DRC) of the total costs, with average indirect costs at 32% of the total costs.

The cost of treatment per young infant covered was lowest in Ile Ife, Nigeria at US$ 0.80 with regimen A compared to US$ 0.70 with regimen D; and highest in Zaria, Nigeria with US$ 1.65 with regimen A and US$ 1.46 with regimen D, mainly because of the higher number of covered young infants as compared to those treated. In 2012, average weighted costs per covered young infant across all sites was estimated at US$ 1.1 (95% CI US$ 0.8–1.3) with regimen A and US$ 0.9 (95% CI US$ 0.7–1.1) with regimen D. For fast breathing, average weighted costs across all sites per covered young infant with regimen E were same as regimen D. In 2018 prices, average weighted costs per covered young infant across all sites were US$ 1.7 for regimen A, US$ 1.5 for regimens B and C; and US$ 1.4 for regimens D and E.

Incremental cost-effectiveness ratios

The incremental cost-effectiveness ratios (ICER) show all three experimental regimens (B, C and D) for CSI and regimen E for fast breathing were cost-effective compared to reference regimen A (Fig 1). The benefits are from both decline in costs and higher effectiveness (that is reduced treatment failure on day 8 after enrollment). For CSI, regimen C is more cost-effective, followed by regimen D as compared to the reference regimen. However, when regimens were compared pairwise, we found that regimen D was more cost-effective as compared to both regimen B and C with ICER at -0.7 and -2.0, respectively for average risk difference in treatment outcomes. For fast breathing, treatment with regimen E was more cost-effective than reference regimen A (ICER = -3.5).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Incremental cost-effectiveness ratios (ICER) for treating young infants with clinical severe infection and fast breathing with different regimens for outcome/effect based on non-treatment failure on day 8 after enrollment.

https://doi.org/10.1371/journal.pone.0247977.g001

Interpretation of results

Our results showed that the cost of treating the CSI sub-category of PSBI in young infants in outpatient settings when the referral to a hospital was not feasible was lowest for regimen D with a combination of an injectable and an oral antibiotic compared to the reference treatment A with only injectables. The cost of regimen D with injection gentamicin once daily for two days plus oral amoxicillin twice daily for seven days was US$ 21 per young infant treated when all costs such as identification, prenatal care, postnatal care, referrals, daily assessments, treatment and indirect administrative and operational costs were taken into account. The weighted average costs for outpatient management of CSI per covered young infant varied between US$ 0.9-US$ 1.1 for any regimen, where coverage was defined as young infants receiving at least one postnatal visit. The ICER for CSI showed all experimental regimens (B, C and D) were more cost-effective, as compared to reference regimen A. WHO guideline recommends using regimen B as option 1 and regimen D as option 2 for treatment of CSI in young infants 0–59 days old when a referral is not feasible [22]. However, pairwise comparisons between regimen B and D; and regimen C and D showed that regimen D was more cost-effective than both regimen B and C. For fast breathing pneumonia alone, treatment with regimen E i.e., oral amoxicillin twice daily for seven days at US$ 18.3 was more cost-effective than reference regimen A at US$ 27.4 and ICER = -3.5. Besides, it can be easily administered by the parent.

The cost per young infant covered and treated varied across sites due to the way interventions are delivered (health center vs. outreach); type of provider used for an intervention; differences in salaries of providers; differences in terrain; variations in costs of consumables and non-consumables; and differences in costs of training and numbers trained. For regimen D, which was found the most cost-effective, the direct costs (Pre-OP plus OP) per young infant treated were US$ 5.4 in the DRC, US$ 12.9 in Kenya and US$ 16.6—US$ 20.4 in Nigeria. The difference was mainly because of the lower salaries of the providers in the DRC as compared to Kenya and Nigeria. However, the cost difference due to lower salary was not proportional, as longer time was taken by the health workers to reach the population in communities due to difficult terrain in the DRC. Cost of medicines and consumables were higher at all sites in Nigeria. Identification of pregnant mothers and sick young infants, prenatal and postnatal visits, linking the sick infant to the facility and assessment of the sick infant were critical for timely identification of PSBI and treatment and constituted almost one-fourth to half of the direct cost. These pre-enrollment and pre-treatment activities costed more in Nigeria as these were undertaken by CHEWs who are more qualified and paid more than the CHWs in other settings. Also, larger numbers of visits to households (per infant treated) had to be made for finding those with danger signs in Nigeria. Daily assessments of treated young infants were considered important for effective outcomes.

Marginal indirect programmatic and administrative costs including management, supervision, meetings between health providers and supervisors, at least one refresher training course annually for all staff delivering services, basic equipment such as weighing scales, thermometers and timers for every CHW/CHEW, communications and travel are important for effective implementation of the program activities for managing PSBI on an outpatient basis when a referral was not feasible. These costs accounted for 32% of the total treatment costs on average, with the highest being for the DRC at 46% and the lowest being in Ibadan, Nigeria at 22% when treated with regimen A.

The total costs (direct and indirect) of managing and treating CSI per young infant treated varied between US$ 32.5 for regimen D to US$ 43.2 for regimen A in 2018 (after adjusting for inflation) and are comparable to those found in Ethiopia (similar setting) for management of PSBI at health posts [24]. The economic costs (including the opportunity costs of the providers) in the Ethiopian study were estimated at US$ 37 (2015 US$), which stated that “adding PSBI management at health post level was estimated to reduce neonatal mortality after day 1 by 17%, translating to a cost per DALY averted of US$ 223 or 47% of GDP per capita, a highly cost-effective intervention by WHO threshold” [24]. In our study, the corresponding average weighted cost per covered young infant was US $1.4 (2018 prices) for regimen D for the average covered population of 8429 young infants. In Ethiopia, costs for management of PSBI at health posts were US$ 1.78 per 100,000 population in a routine setting with 95% of women receiving at least four visits [24].

For fast breathing pneumonia alone, our average outpatient costs per young infant treated across all sites with only oral amoxicillin (regimen E) estimated at US$ 18.3 in 2012 or US$ 28.9 (2018 prices) are comparable to the results for low and middle-income countries (LMIC). For the management of chest indrawing pneumonia in children in LMICs, Zhang and colleagues reported the cost per episode in 2013 US$ was US$ 4.3 in the community, US$ 51.7 in outpatient facilities and US$ 242.7-US$ 559.4 at different levels of hospitals for inpatient settings [23]. Direct medical costs of chest indrawing pneumonia management from Pakistan were reported as US$ 1.5 for community ambulatory care and US$ 7.9 for outpatient care in 2013 US dollars [27], however these don’t include any human resource, programmatic or adminstrative costs. Zhang and colleagues found that the mean length of stay in a hospital for children with chest indrawing pneumonia was 5.8 days in LMICs and 7.7 days in high-income countries.

Manandhar and colleagues argued that an intervention that costs less than US$ 127 was cost-effective [28]. The treatment for PSBI costing less than US$ 43 (2018 prices) with either regimen used in our study at any site was cost-effective by this criterion. Outpatient or community treatment is not only beneficial in terms of reduced costs but is also less disruptive for families and carries less risk of hospital-acquired infections [23]. The reasons for refusal to accept a referral to a hospital include lack of permission from concerned family members, lack of child care, religious and cultural beliefs, distance, cost of travel and treatment, concerns around quality of care and attitudes of health workers [7, 11–13].

Strengths and limitations

Our study’s strengths were that a randomized controlled trial was used to implement the study, along with standardized training of staff and standardized data collection [18]. The costs captured in research settings are normally higher than in routine work/program setting but have been appropriately allocated with robust assumptions for the government program. Some items, such as human resources, programmatic and administrative costs are more difficult to estimate in a government setting, where the providers are engaged in more than one activity. Most studies do not estimate and attribute programmatic and administrative costs for the effective implementation of the program. Our study not only estimated these but also considered pre-treatment costs of actively screening young infants with danger signs and following up with them for treatment and referral. Our direct cost estimates were the economic costs which included the opportunity costs of the providers and depreciated and discounted values of capital. However, only marginal financial costs are used for indirect operational costs. One potential limitation could be that the study estimated the costs data based on the number of young infants treated and covered for 2012. However, we used the inflation factor to estimate the average costs in 2018 prices, which can be used for the advocacy purpose.