Cost of hospital care of women with postpartum haemorrhage in India, Kenya, Nigeria and Uganda: a financial case for improved prevention

Objective: Access to quality, effective lifesaving uterotonics in low and middle-income countries (LMICs) remains a major barrier to reducing maternal deaths from postpartum haemorrhage (PPH). Our objective was to assess the costs of care for women who receive different preventative uterotonics, and with PPH and no-PPH so that the differences, if significant, can inform better resource allocation for maternal health care. Methods: The costs of direct hospital care of women who received oxytocin or heat-stable carbetocin for prevention of PPH in selected tertiary care facilities in India, Kenya, Nigeria, and Uganda were assessed. We collected data from all women who had PPH, as well as a random sample of women without PPH. Cost data was collected for the cost of stay, PPH interventions, transfusions and medications for 2966 women. We analyzed the difference in cost of care at a facility level between women who experienced a PPH event and those who did not. Key findings The mean cost of care of a woman experiencing PPH in the study sites in India, Kenya, Nigeria, and Uganda exceeded the cost of care of a woman who did not experience PPH by between 21% and 309%. There was a large variation in cost across hospitals within a country and across countries. Conclusion: Our results quantify the increased cost of PPH of up to 4.1 times that for a birth without PPH. PPH cost information can help countries to evaluate options across different conditions and in the formulation of appropriate guidelines for intrapartum care, including rational selection of quality-assured, effective medicines. This information can be applied to national assessment and adaptation of international recommendations such as the World Health Organization’s recommendations on uterotonics for the prevention of PPH or other interventions used to treat PPH. Trial registration HRP Trial A65870; UTN U1111-1162-8519; ACTRN12614000870651; CTRI/2016/05/006969, EUDRACT 2014–004445-26. Date of registration 14 August 2014 Plain English summary: Access to quality, effective lifesaving medicines in low and middle-income countries remains a major barrier to reducing maternal deaths from bleeding after childbirth. Information on to what extent treatments for bleeding increases the cost of care of women after childbirth is important for informed resource allocation. We collected data from all women who had bleeding after childbirth, as well as a random sample of women without bleeding in selected hospitals in India, Kenya, Nigeria, and Uganda. Cost data was collected for the cost of stay and interventions to manage bleeding for 2966 women. We compared the difference in cost of care between women who experienced a bleeding event and those who did not. The mean cost of care of a woman with bleeding in the study sites exceeded the cost of care of a woman who did not experience PPH by between 21% and 309%. There was a large variation in cost across hospitals within a country and across countries. Our results indicate an increased cost of bleeding of up to 4.1 times that for birth without bleeding. Effective prevention reduces the cost of care. Cost information can help countries to evaluate options across different conditions and in the formulation of appropriate guidelines for intrapartum care, including rational selection of quality-assured, effective medicines. This information can be applied to national assessment and adaptation of international recommendations such as the World Health Organization’s recommendations on medications for the prevention of bleeding after childbirth or other interventions used to treat bleeding. We calculated the cost of hospital care of women with and without PPH and who received either oxytocin 10 IU or heat-stable carbetocin 100 µg as part of the management of the third stage of labour at nine tertiary referral hospitals (sites) in India, Kenya, Nigeria and Uganda. The CHAMPION trial was a randomized, double-blind, non-inferiority trial comparing the effectiveness in the prevention of PPH of an intramuscular injection of heat-stable carbetocin with oxytocin administered immediately after vaginal birth. The trial methods and results are described in detail elsewhere [10]. Briefly, almost 30,000 women across 23 sites in ten countries were randomly assigned to prophylactic heat-stable carbetocin or oxytocin. The primary outcomes were the proportion of women with blood loss of at least 500 ml or the use of additional uterotonic agents, and the proportion of women with blood loss of at least 1000 ml. Secondary outcomes included the proportion of women having additional interventions to control bleeding. The trial results showed that heat-stable carbetocin was non-inferior to oxytocin. Data from all women who participated in the CHAMPION trial in India, Kenya, Nigeria, and Uganda and who had PPH (n = 1514), as well as a random sample of the same number of women who did not experience PPH (n = 1514), was extracted from the CHAMPION trial database (Fig. 1). One of the ten trial sites was eliminated from this study due to administrative barriers to data collection (n = 62). A total of 2966 records from the CHAMPION trial were therefore included in this study. The four countries were selected because of their high maternal mortality and morbidity rates, climate, and possible difficulties of ensuring the quality of oxytocin due to challenges maintaining the cold chain of oxytocin. Study flow chart The following data, covering the time immediately following administration of the prophylactic uterotonic to the time of discharge from the hospital, were extracted from the CHAMPION trial database to form the basis of the calculation of the cost of care: administration of additional uterotonics; other medical and surgical interventions to treat PPH; blood transfusion and duration of hospital stay. The medical and surgical interventions recorded in the CHAMPION trial patient records and costed in this study were: suturing of the cervix/high vaginal tear; intrauterine balloon/condom tamponade; exploration of the uterine cavity; uterine or hypogastric ligation; manual removal of the placenta; bimanual compression of the uterus and hysterectomy. Uterine compression sutures and manual or surgical correction of uterine inversion were not used for any cases at any of the sites and therefore are not reflected in the results. Excel data collection worksheets were specifically designed for this exercise, modelled on those developed by the Guttmacher Institute for their study estimating the cost of PPH in Egypt [11]. The worksheet design was reviewed by an obstetrician with experience in low income countries and by a local expert panel established in each country by the coordinator of the CHAMPION trial. The correspondent costs were obtained by the CHAMPION trial investigators from each of the participating hospitals. Sites obtained data from various sources including hospital pharmacies, accounting departments, payroll, and government salary data. Table ​Table11 lists all elements included in each of the items costed for the study. Breakdown of costs and details of calculations Cost of disposable apparel (gowns, gloves, caps etc.) + disposable consumables (needles, gauze, syringes etc.) + instruments (sterilization fee) + cleaning + laboratory tests + fee for use of operating theatre (if applicable) + staff labour cost (L) Each surgical intervention was costed as an independent intervention, although some patients may have received more than one intervention concurrently with some costs shared across interventions Monthly salary/hours worked per month x hours devoted to intervention Calculated for all personnel required for the procedure First unit of blood: Cost of apparel + consumables + sterilization charge + laboratory tests + blood product price + cost of labour for blood (LB) Each additional unit of blood: Cost of labour for blood (LB) + blood product price Basic cost of 24 h stay in ward bed x the number of days in hospital from the time of birth to discharge Some patients were discharged within 24 h and with no additional care over basic delivery services. As we did not calculate the cost of delivery specifically (it occurs prior to diagnosis of PPH) these patients’ records would show zero cost. To avoid this, all patients were allocated a minimum of 1 day’s hospital stay as a proxy for the cost of delivery Price of the dose of the uterotonic administered, being either: the price of the individual unit (i.e. tablet, ampoule) multiplied by the units administered or the pack price divided by units in the pack multiplied by the units administered Price of the dose of the medicine administered, being either: the price of the individual unit (i.e. tablet, ampoule) multiplied by the units administered or the pack price divided by units in the pack multiplied by the units administered T2 = Total cost of care HS = Cost of hospital stay I = Costs of a surgical intervention received for treatment of PPH and ΣI = the sum of costs of all surgical interventions received U = Sum of the cost of all additional uterotonics administered for treatment of PPH B = Cost of blood and blood products administered CM = Sum of cost of all concomitant medicines administered HS = Cost of hospital stay I = Costs of a surgical intervention received for treatment of PPH and DT = Cost of the full tray of drugs available for each surgical intervention for PPH treatment U = Sum of the cost of all additional uterotonics administered for treatment of PPH B = Cost of blood and blood products administered The data for concomitant medicines administered during the CHAMPION trial included an extensive list of medicines, many of which were not related to PPH care. It was not possible to isolate the drugs that were administered specifically for the treatment of PPH from the list of concomitant drugs. The cost of the drugs available for each surgical intervention was also collected to complement the cost information. Investigators from each of the countries were requested to list all the drugs used to treat PPH. This list developed per country was named ‘drug tray’ for the purposes of this study The Drug Tray formula was used to validate the results by addressing the uncertainty about the concomitants being related to PPH or not. For the Drug Tray formula, it should be noted that the entirety of the drug tray may not have been used for each intervention depending on each patient’s medical condition. It was not logistically possible to cross-match the concomitants and drug tray data to achieve a more accurate result All prices collected in this study were current at the time of data collection, i.e., between February and July 2018, and not at the time of the CHAMPION trial intervention. The conversion from local currency to US Dollar was made at the time of data collection. Disposable consumables were costed using unit prices provided by the facilities from procurement records. The costs related to the use of non-disposable items, such as surgical instruments, were calculated as the cost of cleaning and sterilization. We did not include depreciation as information on the life of non-disposable items was not available. One on-site meeting with each country investigator was performed by the research team during the data collection period for quality assurance purposes. The total cost of care at the facility for each patient (PPH and no PPH) was computed using three different formulas. These were (1) Total cost of care; (2) Total cost of care including concomitant medicines women received; and (3) Total cost of care including drug tray costs. We did not include the concomitant medicines and drug tray formulas in the main manuscript since, due to variations in the number of medicines used and the contents of the drug trays, the results showed high variability and distracted from the main comparison objective. The concomitant medicines and drug tray formulas are presented in Table ​Table11. The formula used for the total cost of care calculation was: where T1 = Total cost of care, HS = Cost of hospital stay, I = Costs of a surgical intervention received for treatment of PPH and ΣI = the sum of costs of all surgical interventions received, U = Sum of the cost of all additional uterotonics administered for treatment of PPH, B = Cost of blood and blood products administered. The cost of prophylactic uterotonics (oxytocin or heat-stable carbetocin) was not included as their administration occurred prior to diagnosis of PPH. The mean costs of care for women receiving oxytocin or heat-stable carbetocin during the third stage of labour, having PPH or not, were calculated for each site but are not presented in the paper since the two uterotonics were very similar in efficacy.