- Email: [email protected]
Improving blood availability in a district hospital, Bo, Sierra Leone
International Journal of Gynecology & Obstetrics 59 Suppl. 2 Ž1997. S127]S134
Improving blood availability in a district hospital, Bo, Sierra Leone P. Sengeha,U , O. Samai b , S.K. Sidique c , A. Kebbie d , P. Fofanab , S. Stephens c , Ž The Bo PMM Team. a
Di¨ision of Planning, Monitoring and E¨aluation, Ministry of Gender and Children’s Affairs, Freetown, Sierra Leone b Ministry of Health, Southern Pro¨ince, Sierra Leone c Bo Go¨ernment Hospital, Bo, Sierra Leone d Bo Paramedical School, Bo, Sierra Leone
Abstract Preliminary studies: A situation analysis of Bo Government Hospital found there was no functioning blood bank. Focus group discussions revealed negative perceptions of community members regarding donation and transfusion of blood. Inter¨entions: In late 1992, the hospital laboratory was refurbished: a refrigerator, freezer, reagents and other equipment and supplies were provided and technicians were trained. Consequently, small quantities of blood could be stored and rapidly transfused. Other improvements to obstetric services in the hospital were undertaken at the same time. Afterwards, community education to encourage blood donation was done. Results: The number of units of blood drawn increased from 304 to 501 Ž65%. from 1992 to 1993 Žthe first intervention year.; the number actually transfused increased from 296 to 452 Ž53%.. By the third intervention year Ž1995., the annual numbers of units drawn and transfused were 469 and 422, respectively. Case fatality rates for major obstetric complications were lower in all post-intervention years Ž10% in 1993, 7% in 1994, 10% in 1995. as compared to 1992 Ž13%.. Costs: Cost of improvements to blood services in the hospital was nearly US $10 000, of which 77% was used to purchase supplies and equipment. Conclusions: Even in settings with irregular electrical supply, storage of small quantities of blood is possible. Ready availability of blood may contribute to improved quality of obstetric care and improved survival among patients. Q 1997 International Federation of Gynecology and Obstetrics Keywords: Africa; Sierra Leone; Maternal mortality; Obstetric services; Blood; Cost
U
Corresponding author, Ministry of Gender and Children’s Affairs, 9th Floor, Youyi Building, Freetown, Sierra Leone. Tel.: q232 22 240635; fax: q232 22 240803. 0020-7292r97r$17.00 Q 1997 International Federation of Gynecology and Obstetrics PII S0020-7292Ž97. 00157-4
S128
P. Sengeh et al. r International Journal of Gynecology & Obstetrics 59 Suppl. 2 (1997) S127]S134
1. Introduction Sierra Leone is a hilly, fertile country on the west coast of Africa, situated between Guinea and Liberia. Its land area of 72 000 km2 holds an estimated population of over four million w1x. The population is largely rural, with approximately 75% of the people making their living from subsistence farming and mining. Standard health indicators place Sierra Leone among the most disadvantaged of nations. Infant mortality is estimated at 164 deaths per 1000 live births, higher than that in all but two countries in the world w2x. Maternal mortality is estimated at 1800 deaths per 100 000 live births, the highest ratio in the world w3x. With a total fertility rate of 6.5 w2x, a Sierra Leonean woman’s chance of dying as a result of pregnancy is one in seven w3x. Maternal deaths are mainly attributed to obstructed labor, hemorrhage, sepsis, unsafe abortion and eclampsia. Because of the deterioration in economic conditions in the country over the past decade, government and individual expenditures on health have been greatly reduced. This has led to poor quality of health services, including emergency obstetric services necessary to prevent maternal deaths. 1.1. The research area The research area reported on in this document is located in the northern part of Bo District, one of the 12 administrative districts in the country. Estimating from the 1985 census, Bo District currently has a population of approximately 350 000. Districts throughout the country are divided into chiefdoms, traditionally demarcated areas governed by paramount chiefs who wield considerable influence in village decision-making. The research area comprises six of the 15 chiefdoms in the district. The population covered } approximately 53 000 } is not proportionate to the land area because the region is particularly remote and sparsely settled. Most health services in the project area are provided by the Ministry of Health’s primary health care service. There are nine peripheral
health units ŽPHUs. in the study area. They provide curative, preventive and promotive care, including delivery. The patient-to-facility ratio is 5900:1 in the study area; the corresponding figure in the district is 5500:1. The facilities are not evenly distributed around the research area. Six of the nine PHUs are situated at the chiefdom headquarters, rather than at larger population centers. This has left the bulk of the population with long distances to travel to the health facilities. Only 38% of the research area population lives within five km of a PHU. The farthest distance between a village and a PHU is approx. 20 km. The PHUs are staffed by community health officers, endemic disease control unit assistants and maternal and child health aides, who provide midwifery services. Also, there are over 200 trained traditional birth attendants ŽTBAs., numerous untrained TBAs and traditional healers and herbalists in the area. It is estimated that approx. 70% of the approximately 2500 deliveries per year in the project area take place at home and are assisted by TBAs w4x. The Bo District Government Hospital serves as the principal referral point for all problem cases, including those with obstetric complications. The average distance between the referral hospital and the communities in the research area is 56 km, a distance that can sometimes take 2]3 h by vehicle during the rainy season. 1.2. Project purpose Since 1988, the Prevention of Maternal Mortality ŽPMM. Network has been conducting operations research on ways to prevent maternal mortality. The Network consists of 11 teams in West Africa and one team at the Center for Population and Family Health at Columbia University in New York, USA. The nine teams in Ghana and Nigeria are based at universities while the two teams in Sierra Leone are situated within the Ministry of Health. The team in Bo, Sierra Leone, carried out interventions at three levels: in the health facilities, in the communication and referral system and in the communities. This paper describes the
P. Sengeh et al. r International Journal of Gynecology & Obstetrics 59 Suppl. 2 (1997) S127]S134
improvements made in the management of blood at Bo Government Hospital. Other interventions are described elsewhere w5]7x. 2. Preparatory research findings Before planning and implementing the interventions, a preparatory study was conducted in February and March 1990, to determine how obstetric complications were managed in the community, in the peripheral health units and at the hospital itself. Through focus group discussions with community members, TBAs and health staff, and a situation analysis of health facilities, reasons for delay in obtaining care for obstetric complications were identified. The team found that most community members, including pregnant women and untrained TBAs, did not know the causes, or readily recognize the signs and symptoms, of obstetric complications. When a problem was recognized, traditional healers, herbalists or faith healers were often approached for assistance. The time taken by their efforts could delay referral to a PHU or hospital. Moreover, their practices themselves sometimes worsened the woman’s condition. Most villages lacked access to transportation. Roads are very poor, vehicles are few and transportation costs are extremely high, particularly for vehicles hired in emergency situations. Tracking down a vehicle and collecting the money to pay for it often took many hours. Should a woman reach a referral point, it was found that there was often delay in providing prompt treatment. Health facilities lacked surgical capability, emergency materials, sufficient drugs and stored blood. Patients’ families had to purchase these themselves from suppliers in town, further increasing both the delay and cost. 2.1. Findings regarding blood donation and transfusion Barriers specific to the functioning of the blood transfusion system at Bo Government Hospital were identified from both the focus groups and the facility review. There was no functioning refrigerator to store blood for emergency opera-
S129
tions. When patients needed blood, the therapy was based on whole fresh blood transfusion. In most instances, delays occurred as it was very difficult to make immediate identification of appropriate donors for the patient. Even where there were willing donors, other supplies for blood donation and transfusion Že.g. antisera for grouping and cross-matching, blood bags, needles and syringes. often were not available. These items were available in public pharmacies where they were sold to the patients at very high costs. Procuring such items from these sources often caused substantial delay, especially at night. The perception of community members regarding blood donation and transfusion was also a major stumbling block. It was revealed in the focus group discussions that many people feared transfusions because they believed they could acquire sickness, madness or evil behaviors from blood. Thus, while many people were willing to receive blood from relations, they were generally unwilling to get blood from strangers. Many people were reluctant to donate blood, even to relatives, because they feared falling sick or they believed that they did not have enough blood to donate. On many occasions, only women accompanied patients to the hospital. This presented a problem since women are traditionally prevented from donating blood and health care providers prefer not to bleed women, who are more likely than men to be anemic. Other beliefs based on religious taboos and blood transfusion were cited as additional constraining factors. 3. Project interventions As noted above, interventions to address the identified problems were carried out at three levels: in the health facilities, including the hospital; in the communication and referral system; and in the communities. Interventions at the health facilities focused on upgrading the physical structures and improving the services Žincluding provision of the necessary drugs and equipment and training of health staff to provide emergency obstetric care.. In order to improve the referral system, equipment for
S130
P. Sengeh et al. r International Journal of Gynecology & Obstetrics 59 Suppl. 2 (1997) S127]S134
transportation and communication } emergency referral vehicle, motorbikes and mobile radios } was provided. At the community level, a continuing health education program was designed to improve community members’ knowledge, attitudes and practices regarding the management of obstetric complications. Health messages were designed by health staff and were disseminated by them and by some members of the community including traditional birth attendants. Within Bo Government Hospital, improvements were made in the laboratory and ablood bank was provided. All the blood-related interventions Ždescribed below. were carried out from July to November 1992. 3.1. Reno¨ation and enlargement of the laboratory A major change instituted in Bo Government Hospital by the team was the creation of a functioning blood bank. First, the existing laboratory was renovated and extended to accommodate blood donors. A bed to serve as a lying-in bed for donors was provided by the hospital management. 3.2. Pro¨ision of equipment and supplies With the assistance of the laboratory superintendent, the necessary equipment and supplies for the blood bank were identified and purchased from the International Drug Administration in Amsterdam. Equipment and supplies } including a freezer, a kerosene-capable refrigerator, sera, blood bags, reagents, needles and syringes } were purchased. 3.3. Procurement of blood bank refrigerator In order to make the blood bank fully functional, an appeal was made to the Department of Health to provide the hospital with an unused but fairly new and fully functional blood bank refrigerator that had been discarded in the National Referral Hospital in Freetown, the country’s capital. This refrigerator had been out of use for over 10 years. With the consent of the chief medical officer, the refrigerator was taken to Bo by the team, cleaned and put into use. With continuous
electricity flow, the refrigerator could now be used to store blood at the optimum temperature of 48C for up to 21 days. This refrigerator was used mainly for the storage of blood. The freezer and kerosene-capable refrigerator Žfrom Amsterdam. were very useful at times of electricity power cuts, particularly during the dry season. 3.4. Staff training A two-day workshop was conducted for the hospital staff who were directly involved in the processing and administering of blood to patients. These included laboratory staff and maternal and child health aides. The workshop’s specific objectives were to increase staff members’ knowledge regarding the processing and administering of whole blood and packed cells and to help the ward staff minimize risks in the management of blood transfusion therapy. The training was conducted by senior hospital staff, the PMM team’s nurse-midwife and a Red Cross worker. Guidelines were developed in the workshop for ordering transfusions, handling blood and recruiting donors. It was decided that these guidelines would be mandatory for all staff. Since the laboratory superintendent was trained to screen for HIV seropositivity, it was also decided that all donors’ blood would be tested. 3.5. Mobilization of Red Cross ¨olunteers The team worked with local Red Cross volunteers who were available to give blood for emergency operations in the hospital. The Red Cross volunteers did not demand money or payment of any kind from patients. During the project, the existing system of calling on the volunteers as emergencies arose was changed. The new system relied on keeping a supply of at least three units of type ‘O’ blood on hand at all times to assure adequate blood for at least two emergencies. The immediate transfusion of type ‘O’ blood to a patient provided the opportunity to contact the patient’s family members or the Red Cross volunteers to obtain supplies of the patient’s specific blood type should additional transfusion be needed. Relatives of patients were strongly urged
P. Sengeh et al. r International Journal of Gynecology & Obstetrics 59 Suppl. 2 (1997) S127]S134
to donate blood to replenish the stored units their family members had used, which they did. 3.6. Community education The Bo team mounted a community education campaign to dispel the negative perceptions about blood donation and transfusion and to explain the benefits of this therapy. In village meetings, team members and elders decided that four people should accompany each woman with an obstetric complication to the hospital. They would serve as potential blood donors, in case a cesarean section or other emergency required it. 4. Costs of the intervention Using data from the project’s cost-tracking records, the sources and amounts of funds used to establish the intervention were analyzed Žsee Table 1.. The cost of the intervention Žexcluding salaries . was approx. US $10 000. The majority of the funds were spent for equipment and supplies. In addition to the cost of the blood bank refrigerator Ž$2000., major equipment costs included the following Žnot shown.: a kerosenerelectric refrigerator Ž$1750.; a freezer Ž$1200.; and laboratory equipment Ž$1670.. Supplies included such consumables as antisera and blood bags Ž$1019, not shown.. Most of the funds Ž78%. were supplied by the Carnegie Corporation, through the PMM project, while the government contributed 21% of the
S131
costs and the community contributed 1%. It should be noted that many expenses that were not included in these calculations } such as staff salaries and infrastructure } were also covered by the government. 5. Data sources and methods The team measured the immediate and more distant effects of establishing the blood bank in the hospital. The most direct measure of the functioning of the blood bank was the number of units of blood drawn and transfused. These data were collected from record books kept in the blood bank. Improving blood availability was not the ultimate purpose of the project, however. It was expected that a well-functioning blood bank would contribute to improved quality of obstetric Žand other. care. The obstetric case fatality rate ŽCFR; the likelihood of death among women admitted with obstetric complications. was used as an indicator of quality of care. Data on numbers of maternal deaths and women admitted with complications from 1992 to 1995 were collected from maternity ward admissions registers. It should be noted that the blood supply is just one of several factors that might affect the hospital’s ability to successfully treat women with obstetric complications. Within this PMM project, other interventions were designed to affect outcome. These included improvements in other aspects of emergency treatment in the hospital and activities to reduce delay in getting women to the
Table 1 Costs and sources of funds, blood bank intervention, Bo Government Hospital, Sierra Leone, 1992}1995 Cost component
Cost ŽUS $.
Percent of total Ž%.
Contributor Žpercent of component. Project Ž%.
Government Ž%.
Community Ž%.
Supplies and equipment Blood bank refrigerator Renovation of laboratory building Training workshop Community education Blood donor mobilization
5639 2000 1594 376 200 60
57 20 16 4 2 -1
98 0 100 70 100 100
2 100 0 5 0 0
0 0 0 25 0 0
Total
9869
100
78
21
1
S132
P. Sengeh et al. r International Journal of Gynecology & Obstetrics 59 Suppl. 2 (1997) S127]S134
hospital by increasing the public’s awareness of complications and providing emergency transport. In addition, political, social and economic events in Sierra Leone, over which the project had no control, clearly affected people’s behavior. It also affected the ability of all systems, both government and private, to function. The project area was close to the zone where government and rebel forces clashed, further exacerbating the uncertainty. Disruption was particularly severe in 1994. 6. Results 6.1. Quantity of blood drawn and transfused There was an overall increase in the number of units of blood drawn annually from the pre-intervention years Ž1990]1992. to the post-intervention years Ž1993]1995.. As shown in Table 2, 469 units were drawn in 1995, as compared with 320 units in 1990 Ža 47% increase .. The number of units actually transfused also rose, from 304 in 1990 to 422 in 1995, an increase of 39%. In the pre-intervention period, almost all units of blood drawn were actually transfused, since blood was drawn after the need was identified in a specific patient. The proportion of units actually transfused decreased somewhat in the post-intervention period, since the hospital now had a stored blood system. Under this system, 90% of the blood was used, as compared to 95]98% under the earlier system.
There was a marked increase Ž80%. in the number of units used for obstetric patients, from 119 units in 1990 to 214 units in 1995. The proportion of transfused blood used for obstetric patients also increased, from 39% in 1990 to 51% in 1995. It is evident from the data that 1994 was a problematic year. The decrease in blood bank activity, as compared to the preceding and the following years, reflects changes in the numbers of women with complications coming to the hospital, rather than a problem in the availability of blood itself. The decline in hospital utilization in 1994 is most likely accounted for by the disruption caused by the rebel war. It is interesting to note that obstetric transfusions as a proportion of all transfusions increased from 1993 to 1994, even though the actual number of these transfusions declined. This suggests that utilization of the hospital for obstetric emergencies declined less than did utilization for other types of emergencies during this difficult year. 6.2. Case fatality rates Case fatality was 13% at Bo Government Hospital in 1992, the year the blood bank system was improved. ŽData prior to 1992 are not available.. The CFR fell to a low of 7% in 1994, but increased to 10% in 1995. While the overall decline is encouraging, a CFR of 10% is still very high. It should be noted that the number of women with complications admitted to the hospital declined during this period Žsee Table 3..
Table 2 Utilization of blood, Bo Government Hospital, Sierra Leone, 1990]1995 Pre-intervention period
Units of blood drawn Units transfused Number Percent of units drawn Units transfused to obstetric patients Number Percent of transfused units
Post-intervention period
1990
1991
1992
1993
1994
1995
320
365
304
501
340
469
304 95%
358 98%
296 97%
452 90%
308 91%
422 90%
119 39%
132 37%
133 45%
235 52%
191 62%
214 51%
P. Sengeh et al. r International Journal of Gynecology & Obstetrics 59 Suppl. 2 (1997) S127]S134 Table 3 Obstetric case fatality rates, Bo Government Hospital, Sierra Leone, 1992]1995 1992 All obstetric patients Maternal deaths Women with complications CFR Patients from project areasU Maternal deaths Women with complications CFR
1993
1994
1995
28 19 11 17 223 184 160 171 13% 10% 7% 10% 1 } 25 } 4% }
} } }
Patients from non-project areas Maternal deaths 26 18 } Women with complications 206 159 } CFR 13% 11% }
} } }
2 17 12%
U
CFRs for patients from project and non-project areas were not calculated for 1994 and 1995 due to the displacement of the population by the rebel war.
Among women with complications from project areas, case fatality rate declined from 12% to 4% from 1992 to 1993, although the numbers are too small to be conclusive. By contrast, among women with complications from non-project areas, case fatality rate remained relatively stable, during the same time period, going from 13% in 1992 to 11% in 1993. ŽCFRs for projectrnon-project areas were not calculated for 1994 and 1995 because rebel activities had intensified, displacing much of the population.. 7. Discussion A component of emergency obstetric care is the availability of blood transfusion to manage complications. The data show an increase in the number of units of blood transfused at Bo Government Hospital after the upgrading of the blood bank in November 1992. The improved blood bank system was particularly beneficial for the maternity ward, as evidenced by the increase in number of units transfused to obstetric patients. Approximately half of all blood transfusions were to non-obstetric patients, illustrating that the project aided other services as well.
S133
The percentage of units wasted increased slightly following the intervention, compared to the pre-intervention period, when the blood bank lacked storage capabilities. However, the 10% wastage in the post-intervention period is an acceptable cost for immediate access to stored blood. An attempt was made by the PMM team to measure the time interval between the request for blood and the actual transfusion. The attempt was unsuccessful, as data were not reliably recorded. Based on observation and interviews, however, the researchers believe there was indeed a reduction in delay. Under the pre-intervention system, a donor had to be identified, typed and bled before a transfusion could take place. This process usually took 2]4 h. Under the stored blood system, the interval was as little as 10 min, or the time it took to walk from the ward to the laboratory to retrieve the stored blood. The case fatality rate declined from 13% in 1992 to 10% in 1995, and the improved blood bank system may have contributed. However, the difference in case fatality between project and non-project area patients suggests that the presence of the blood bank Žto which all patients had equal access. might have been less important than some other factors Žwhich were only available to project-area residents., although again, the numbers are too small to be conclusive. For example, women from the project area were entitled to hospital treatment before payment, whereas those from elsewhere still had to pay before being treated. The difference in CFR may also reflect the effects of other project interventions in the project area, such as improved services and referrals at PHUs, the emergency transportation system, community health education, and access to community loan funds in some communities. We consider the cost of the blood bank intervention, approximately $10 000, to be a reasonable expenditure to permit a district hospital to have blood storage capability. This project has demonstrated that a blood storage system, for a small number of units, is feasible even in the context of irregular electrical supply. The improvements in the management of
S134
P. Sengeh et al. r International Journal of Gynecology & Obstetrics 59 Suppl. 2 (1997) S127]S134
the blood bank system are clear examples of how discarded resources can be used and how communities and agencies such as the Red Cross can be involved to improve the health status of the people. These improvements are not only felt in the six project-area chiefdoms, but in the entire district that Bo Government Hospital serves. The replicability of research efforts is often an issue of concern to program managers. Several characteristics of the PMM project in Bo District reinforce its replicability. While the project had financial and technical assistance from outside sources, many of the improvements were made possible through the creative use of local resources. An objective inventory of space and equipment identified an unused obstetric theater which needed only minor refurbishing, and a functioning blood bank in storage. While this was perhaps especially fortunate, it is not uncommon to find under-used space, material and equipment in many health facilities. Acknowledgements We are extremely grateful to the Carnegie Corporation of New York for providing funds to undertake this research. We are also thankful to the Columbia University PMM Network team for the extensive technical assistance given to us. To the other team members of the PMM Network in
West Africa, we say many thanks for the numerous experiences we shared during the six years of our working together. Finally, we say many thanks to our benefactors, the local Red Cross blood donors’ club in Bo Town and many other people who willingly donated their blood to save many lives. References w1x Central Statistics Office. Provisional results: 1985 census. Freetown, Sierra Leone: Central Statistics Office, 1987. w2x UNICEF. State of the world’s children, 1995. New York: Oxford University Press, 1995. w3x World Health OrganizationrUNICEF. Revised 1990 estimates of maternal mortality: a new approach by WHO a n d U N I C E F . W H O r F R H r M S M r 9 6 .1 1 ; UNICEFrPLNr96.1, April 1996. w4x BorPujehun Primary Health Care Programme Monitoring and Evaluation Unit. Prospective infant and maternal mortality survey results. Sierra Leone, 1990 Žunpublished.. w5x Thuray H, Samai O, Fofana P, Sengeh P. Establishing a cost recovery system for drugs, Bo, Sierra Leone. Int J Gynecol Obstet 1997;59ŽSuppl. 2.:141]147. w6x Fofana P, Samai O, Kebbie A, Sengeh P. Promoting use of obstetric services through community loan funds, Bo, Sierra Leone. Int J Gynecol Obstet 1997;59ŽSuppl. 2.:225]230. w7x Samai O, Sengeh P. Facilitating emergency obstetric care through transportation and communication, Bo, Sierra Leone. Int J Gynecol Obstet 1997;59ŽSuppl. 2.:157]164.
Improving blood availability in a district hospital, Bo, Sierra Leone P. Sengeha,U , O. Samai b , S.K. Sidique c , A. Kebbie d , P. Fofanab , S. Stephens c , Ž The Bo PMM Team. a
Di¨ision of Planning, Monitoring and E¨aluation, Ministry of Gender and Children’s Affairs, Freetown, Sierra Leone b Ministry of Health, Southern Pro¨ince, Sierra Leone c Bo Go¨ernment Hospital, Bo, Sierra Leone d Bo Paramedical School, Bo, Sierra Leone
Abstract Preliminary studies: A situation analysis of Bo Government Hospital found there was no functioning blood bank. Focus group discussions revealed negative perceptions of community members regarding donation and transfusion of blood. Inter¨entions: In late 1992, the hospital laboratory was refurbished: a refrigerator, freezer, reagents and other equipment and supplies were provided and technicians were trained. Consequently, small quantities of blood could be stored and rapidly transfused. Other improvements to obstetric services in the hospital were undertaken at the same time. Afterwards, community education to encourage blood donation was done. Results: The number of units of blood drawn increased from 304 to 501 Ž65%. from 1992 to 1993 Žthe first intervention year.; the number actually transfused increased from 296 to 452 Ž53%.. By the third intervention year Ž1995., the annual numbers of units drawn and transfused were 469 and 422, respectively. Case fatality rates for major obstetric complications were lower in all post-intervention years Ž10% in 1993, 7% in 1994, 10% in 1995. as compared to 1992 Ž13%.. Costs: Cost of improvements to blood services in the hospital was nearly US $10 000, of which 77% was used to purchase supplies and equipment. Conclusions: Even in settings with irregular electrical supply, storage of small quantities of blood is possible. Ready availability of blood may contribute to improved quality of obstetric care and improved survival among patients. Q 1997 International Federation of Gynecology and Obstetrics Keywords: Africa; Sierra Leone; Maternal mortality; Obstetric services; Blood; Cost
U
Corresponding author, Ministry of Gender and Children’s Affairs, 9th Floor, Youyi Building, Freetown, Sierra Leone. Tel.: q232 22 240635; fax: q232 22 240803. 0020-7292r97r$17.00 Q 1997 International Federation of Gynecology and Obstetrics PII S0020-7292Ž97. 00157-4
S128
P. Sengeh et al. r International Journal of Gynecology & Obstetrics 59 Suppl. 2 (1997) S127]S134
1. Introduction Sierra Leone is a hilly, fertile country on the west coast of Africa, situated between Guinea and Liberia. Its land area of 72 000 km2 holds an estimated population of over four million w1x. The population is largely rural, with approximately 75% of the people making their living from subsistence farming and mining. Standard health indicators place Sierra Leone among the most disadvantaged of nations. Infant mortality is estimated at 164 deaths per 1000 live births, higher than that in all but two countries in the world w2x. Maternal mortality is estimated at 1800 deaths per 100 000 live births, the highest ratio in the world w3x. With a total fertility rate of 6.5 w2x, a Sierra Leonean woman’s chance of dying as a result of pregnancy is one in seven w3x. Maternal deaths are mainly attributed to obstructed labor, hemorrhage, sepsis, unsafe abortion and eclampsia. Because of the deterioration in economic conditions in the country over the past decade, government and individual expenditures on health have been greatly reduced. This has led to poor quality of health services, including emergency obstetric services necessary to prevent maternal deaths. 1.1. The research area The research area reported on in this document is located in the northern part of Bo District, one of the 12 administrative districts in the country. Estimating from the 1985 census, Bo District currently has a population of approximately 350 000. Districts throughout the country are divided into chiefdoms, traditionally demarcated areas governed by paramount chiefs who wield considerable influence in village decision-making. The research area comprises six of the 15 chiefdoms in the district. The population covered } approximately 53 000 } is not proportionate to the land area because the region is particularly remote and sparsely settled. Most health services in the project area are provided by the Ministry of Health’s primary health care service. There are nine peripheral
health units ŽPHUs. in the study area. They provide curative, preventive and promotive care, including delivery. The patient-to-facility ratio is 5900:1 in the study area; the corresponding figure in the district is 5500:1. The facilities are not evenly distributed around the research area. Six of the nine PHUs are situated at the chiefdom headquarters, rather than at larger population centers. This has left the bulk of the population with long distances to travel to the health facilities. Only 38% of the research area population lives within five km of a PHU. The farthest distance between a village and a PHU is approx. 20 km. The PHUs are staffed by community health officers, endemic disease control unit assistants and maternal and child health aides, who provide midwifery services. Also, there are over 200 trained traditional birth attendants ŽTBAs., numerous untrained TBAs and traditional healers and herbalists in the area. It is estimated that approx. 70% of the approximately 2500 deliveries per year in the project area take place at home and are assisted by TBAs w4x. The Bo District Government Hospital serves as the principal referral point for all problem cases, including those with obstetric complications. The average distance between the referral hospital and the communities in the research area is 56 km, a distance that can sometimes take 2]3 h by vehicle during the rainy season. 1.2. Project purpose Since 1988, the Prevention of Maternal Mortality ŽPMM. Network has been conducting operations research on ways to prevent maternal mortality. The Network consists of 11 teams in West Africa and one team at the Center for Population and Family Health at Columbia University in New York, USA. The nine teams in Ghana and Nigeria are based at universities while the two teams in Sierra Leone are situated within the Ministry of Health. The team in Bo, Sierra Leone, carried out interventions at three levels: in the health facilities, in the communication and referral system and in the communities. This paper describes the
P. Sengeh et al. r International Journal of Gynecology & Obstetrics 59 Suppl. 2 (1997) S127]S134
improvements made in the management of blood at Bo Government Hospital. Other interventions are described elsewhere w5]7x. 2. Preparatory research findings Before planning and implementing the interventions, a preparatory study was conducted in February and March 1990, to determine how obstetric complications were managed in the community, in the peripheral health units and at the hospital itself. Through focus group discussions with community members, TBAs and health staff, and a situation analysis of health facilities, reasons for delay in obtaining care for obstetric complications were identified. The team found that most community members, including pregnant women and untrained TBAs, did not know the causes, or readily recognize the signs and symptoms, of obstetric complications. When a problem was recognized, traditional healers, herbalists or faith healers were often approached for assistance. The time taken by their efforts could delay referral to a PHU or hospital. Moreover, their practices themselves sometimes worsened the woman’s condition. Most villages lacked access to transportation. Roads are very poor, vehicles are few and transportation costs are extremely high, particularly for vehicles hired in emergency situations. Tracking down a vehicle and collecting the money to pay for it often took many hours. Should a woman reach a referral point, it was found that there was often delay in providing prompt treatment. Health facilities lacked surgical capability, emergency materials, sufficient drugs and stored blood. Patients’ families had to purchase these themselves from suppliers in town, further increasing both the delay and cost. 2.1. Findings regarding blood donation and transfusion Barriers specific to the functioning of the blood transfusion system at Bo Government Hospital were identified from both the focus groups and the facility review. There was no functioning refrigerator to store blood for emergency opera-
S129
tions. When patients needed blood, the therapy was based on whole fresh blood transfusion. In most instances, delays occurred as it was very difficult to make immediate identification of appropriate donors for the patient. Even where there were willing donors, other supplies for blood donation and transfusion Že.g. antisera for grouping and cross-matching, blood bags, needles and syringes. often were not available. These items were available in public pharmacies where they were sold to the patients at very high costs. Procuring such items from these sources often caused substantial delay, especially at night. The perception of community members regarding blood donation and transfusion was also a major stumbling block. It was revealed in the focus group discussions that many people feared transfusions because they believed they could acquire sickness, madness or evil behaviors from blood. Thus, while many people were willing to receive blood from relations, they were generally unwilling to get blood from strangers. Many people were reluctant to donate blood, even to relatives, because they feared falling sick or they believed that they did not have enough blood to donate. On many occasions, only women accompanied patients to the hospital. This presented a problem since women are traditionally prevented from donating blood and health care providers prefer not to bleed women, who are more likely than men to be anemic. Other beliefs based on religious taboos and blood transfusion were cited as additional constraining factors. 3. Project interventions As noted above, interventions to address the identified problems were carried out at three levels: in the health facilities, including the hospital; in the communication and referral system; and in the communities. Interventions at the health facilities focused on upgrading the physical structures and improving the services Žincluding provision of the necessary drugs and equipment and training of health staff to provide emergency obstetric care.. In order to improve the referral system, equipment for
S130
P. Sengeh et al. r International Journal of Gynecology & Obstetrics 59 Suppl. 2 (1997) S127]S134
transportation and communication } emergency referral vehicle, motorbikes and mobile radios } was provided. At the community level, a continuing health education program was designed to improve community members’ knowledge, attitudes and practices regarding the management of obstetric complications. Health messages were designed by health staff and were disseminated by them and by some members of the community including traditional birth attendants. Within Bo Government Hospital, improvements were made in the laboratory and ablood bank was provided. All the blood-related interventions Ždescribed below. were carried out from July to November 1992. 3.1. Reno¨ation and enlargement of the laboratory A major change instituted in Bo Government Hospital by the team was the creation of a functioning blood bank. First, the existing laboratory was renovated and extended to accommodate blood donors. A bed to serve as a lying-in bed for donors was provided by the hospital management. 3.2. Pro¨ision of equipment and supplies With the assistance of the laboratory superintendent, the necessary equipment and supplies for the blood bank were identified and purchased from the International Drug Administration in Amsterdam. Equipment and supplies } including a freezer, a kerosene-capable refrigerator, sera, blood bags, reagents, needles and syringes } were purchased. 3.3. Procurement of blood bank refrigerator In order to make the blood bank fully functional, an appeal was made to the Department of Health to provide the hospital with an unused but fairly new and fully functional blood bank refrigerator that had been discarded in the National Referral Hospital in Freetown, the country’s capital. This refrigerator had been out of use for over 10 years. With the consent of the chief medical officer, the refrigerator was taken to Bo by the team, cleaned and put into use. With continuous
electricity flow, the refrigerator could now be used to store blood at the optimum temperature of 48C for up to 21 days. This refrigerator was used mainly for the storage of blood. The freezer and kerosene-capable refrigerator Žfrom Amsterdam. were very useful at times of electricity power cuts, particularly during the dry season. 3.4. Staff training A two-day workshop was conducted for the hospital staff who were directly involved in the processing and administering of blood to patients. These included laboratory staff and maternal and child health aides. The workshop’s specific objectives were to increase staff members’ knowledge regarding the processing and administering of whole blood and packed cells and to help the ward staff minimize risks in the management of blood transfusion therapy. The training was conducted by senior hospital staff, the PMM team’s nurse-midwife and a Red Cross worker. Guidelines were developed in the workshop for ordering transfusions, handling blood and recruiting donors. It was decided that these guidelines would be mandatory for all staff. Since the laboratory superintendent was trained to screen for HIV seropositivity, it was also decided that all donors’ blood would be tested. 3.5. Mobilization of Red Cross ¨olunteers The team worked with local Red Cross volunteers who were available to give blood for emergency operations in the hospital. The Red Cross volunteers did not demand money or payment of any kind from patients. During the project, the existing system of calling on the volunteers as emergencies arose was changed. The new system relied on keeping a supply of at least three units of type ‘O’ blood on hand at all times to assure adequate blood for at least two emergencies. The immediate transfusion of type ‘O’ blood to a patient provided the opportunity to contact the patient’s family members or the Red Cross volunteers to obtain supplies of the patient’s specific blood type should additional transfusion be needed. Relatives of patients were strongly urged
P. Sengeh et al. r International Journal of Gynecology & Obstetrics 59 Suppl. 2 (1997) S127]S134
to donate blood to replenish the stored units their family members had used, which they did. 3.6. Community education The Bo team mounted a community education campaign to dispel the negative perceptions about blood donation and transfusion and to explain the benefits of this therapy. In village meetings, team members and elders decided that four people should accompany each woman with an obstetric complication to the hospital. They would serve as potential blood donors, in case a cesarean section or other emergency required it. 4. Costs of the intervention Using data from the project’s cost-tracking records, the sources and amounts of funds used to establish the intervention were analyzed Žsee Table 1.. The cost of the intervention Žexcluding salaries . was approx. US $10 000. The majority of the funds were spent for equipment and supplies. In addition to the cost of the blood bank refrigerator Ž$2000., major equipment costs included the following Žnot shown.: a kerosenerelectric refrigerator Ž$1750.; a freezer Ž$1200.; and laboratory equipment Ž$1670.. Supplies included such consumables as antisera and blood bags Ž$1019, not shown.. Most of the funds Ž78%. were supplied by the Carnegie Corporation, through the PMM project, while the government contributed 21% of the
S131
costs and the community contributed 1%. It should be noted that many expenses that were not included in these calculations } such as staff salaries and infrastructure } were also covered by the government. 5. Data sources and methods The team measured the immediate and more distant effects of establishing the blood bank in the hospital. The most direct measure of the functioning of the blood bank was the number of units of blood drawn and transfused. These data were collected from record books kept in the blood bank. Improving blood availability was not the ultimate purpose of the project, however. It was expected that a well-functioning blood bank would contribute to improved quality of obstetric Žand other. care. The obstetric case fatality rate ŽCFR; the likelihood of death among women admitted with obstetric complications. was used as an indicator of quality of care. Data on numbers of maternal deaths and women admitted with complications from 1992 to 1995 were collected from maternity ward admissions registers. It should be noted that the blood supply is just one of several factors that might affect the hospital’s ability to successfully treat women with obstetric complications. Within this PMM project, other interventions were designed to affect outcome. These included improvements in other aspects of emergency treatment in the hospital and activities to reduce delay in getting women to the
Table 1 Costs and sources of funds, blood bank intervention, Bo Government Hospital, Sierra Leone, 1992}1995 Cost component
Cost ŽUS $.
Percent of total Ž%.
Contributor Žpercent of component. Project Ž%.
Government Ž%.
Community Ž%.
Supplies and equipment Blood bank refrigerator Renovation of laboratory building Training workshop Community education Blood donor mobilization
5639 2000 1594 376 200 60
57 20 16 4 2 -1
98 0 100 70 100 100
2 100 0 5 0 0
0 0 0 25 0 0
Total
9869
100
78
21
1
S132
P. Sengeh et al. r International Journal of Gynecology & Obstetrics 59 Suppl. 2 (1997) S127]S134
hospital by increasing the public’s awareness of complications and providing emergency transport. In addition, political, social and economic events in Sierra Leone, over which the project had no control, clearly affected people’s behavior. It also affected the ability of all systems, both government and private, to function. The project area was close to the zone where government and rebel forces clashed, further exacerbating the uncertainty. Disruption was particularly severe in 1994. 6. Results 6.1. Quantity of blood drawn and transfused There was an overall increase in the number of units of blood drawn annually from the pre-intervention years Ž1990]1992. to the post-intervention years Ž1993]1995.. As shown in Table 2, 469 units were drawn in 1995, as compared with 320 units in 1990 Ža 47% increase .. The number of units actually transfused also rose, from 304 in 1990 to 422 in 1995, an increase of 39%. In the pre-intervention period, almost all units of blood drawn were actually transfused, since blood was drawn after the need was identified in a specific patient. The proportion of units actually transfused decreased somewhat in the post-intervention period, since the hospital now had a stored blood system. Under this system, 90% of the blood was used, as compared to 95]98% under the earlier system.
There was a marked increase Ž80%. in the number of units used for obstetric patients, from 119 units in 1990 to 214 units in 1995. The proportion of transfused blood used for obstetric patients also increased, from 39% in 1990 to 51% in 1995. It is evident from the data that 1994 was a problematic year. The decrease in blood bank activity, as compared to the preceding and the following years, reflects changes in the numbers of women with complications coming to the hospital, rather than a problem in the availability of blood itself. The decline in hospital utilization in 1994 is most likely accounted for by the disruption caused by the rebel war. It is interesting to note that obstetric transfusions as a proportion of all transfusions increased from 1993 to 1994, even though the actual number of these transfusions declined. This suggests that utilization of the hospital for obstetric emergencies declined less than did utilization for other types of emergencies during this difficult year. 6.2. Case fatality rates Case fatality was 13% at Bo Government Hospital in 1992, the year the blood bank system was improved. ŽData prior to 1992 are not available.. The CFR fell to a low of 7% in 1994, but increased to 10% in 1995. While the overall decline is encouraging, a CFR of 10% is still very high. It should be noted that the number of women with complications admitted to the hospital declined during this period Žsee Table 3..
Table 2 Utilization of blood, Bo Government Hospital, Sierra Leone, 1990]1995 Pre-intervention period
Units of blood drawn Units transfused Number Percent of units drawn Units transfused to obstetric patients Number Percent of transfused units
Post-intervention period
1990
1991
1992
1993
1994
1995
320
365
304
501
340
469
304 95%
358 98%
296 97%
452 90%
308 91%
422 90%
119 39%
132 37%
133 45%
235 52%
191 62%
214 51%
P. Sengeh et al. r International Journal of Gynecology & Obstetrics 59 Suppl. 2 (1997) S127]S134 Table 3 Obstetric case fatality rates, Bo Government Hospital, Sierra Leone, 1992]1995 1992 All obstetric patients Maternal deaths Women with complications CFR Patients from project areasU Maternal deaths Women with complications CFR
1993
1994
1995
28 19 11 17 223 184 160 171 13% 10% 7% 10% 1 } 25 } 4% }
} } }
Patients from non-project areas Maternal deaths 26 18 } Women with complications 206 159 } CFR 13% 11% }
} } }
2 17 12%
U
CFRs for patients from project and non-project areas were not calculated for 1994 and 1995 due to the displacement of the population by the rebel war.
Among women with complications from project areas, case fatality rate declined from 12% to 4% from 1992 to 1993, although the numbers are too small to be conclusive. By contrast, among women with complications from non-project areas, case fatality rate remained relatively stable, during the same time period, going from 13% in 1992 to 11% in 1993. ŽCFRs for projectrnon-project areas were not calculated for 1994 and 1995 because rebel activities had intensified, displacing much of the population.. 7. Discussion A component of emergency obstetric care is the availability of blood transfusion to manage complications. The data show an increase in the number of units of blood transfused at Bo Government Hospital after the upgrading of the blood bank in November 1992. The improved blood bank system was particularly beneficial for the maternity ward, as evidenced by the increase in number of units transfused to obstetric patients. Approximately half of all blood transfusions were to non-obstetric patients, illustrating that the project aided other services as well.
S133
The percentage of units wasted increased slightly following the intervention, compared to the pre-intervention period, when the blood bank lacked storage capabilities. However, the 10% wastage in the post-intervention period is an acceptable cost for immediate access to stored blood. An attempt was made by the PMM team to measure the time interval between the request for blood and the actual transfusion. The attempt was unsuccessful, as data were not reliably recorded. Based on observation and interviews, however, the researchers believe there was indeed a reduction in delay. Under the pre-intervention system, a donor had to be identified, typed and bled before a transfusion could take place. This process usually took 2]4 h. Under the stored blood system, the interval was as little as 10 min, or the time it took to walk from the ward to the laboratory to retrieve the stored blood. The case fatality rate declined from 13% in 1992 to 10% in 1995, and the improved blood bank system may have contributed. However, the difference in case fatality between project and non-project area patients suggests that the presence of the blood bank Žto which all patients had equal access. might have been less important than some other factors Žwhich were only available to project-area residents., although again, the numbers are too small to be conclusive. For example, women from the project area were entitled to hospital treatment before payment, whereas those from elsewhere still had to pay before being treated. The difference in CFR may also reflect the effects of other project interventions in the project area, such as improved services and referrals at PHUs, the emergency transportation system, community health education, and access to community loan funds in some communities. We consider the cost of the blood bank intervention, approximately $10 000, to be a reasonable expenditure to permit a district hospital to have blood storage capability. This project has demonstrated that a blood storage system, for a small number of units, is feasible even in the context of irregular electrical supply. The improvements in the management of
S134
P. Sengeh et al. r International Journal of Gynecology & Obstetrics 59 Suppl. 2 (1997) S127]S134
the blood bank system are clear examples of how discarded resources can be used and how communities and agencies such as the Red Cross can be involved to improve the health status of the people. These improvements are not only felt in the six project-area chiefdoms, but in the entire district that Bo Government Hospital serves. The replicability of research efforts is often an issue of concern to program managers. Several characteristics of the PMM project in Bo District reinforce its replicability. While the project had financial and technical assistance from outside sources, many of the improvements were made possible through the creative use of local resources. An objective inventory of space and equipment identified an unused obstetric theater which needed only minor refurbishing, and a functioning blood bank in storage. While this was perhaps especially fortunate, it is not uncommon to find under-used space, material and equipment in many health facilities. Acknowledgements We are extremely grateful to the Carnegie Corporation of New York for providing funds to undertake this research. We are also thankful to the Columbia University PMM Network team for the extensive technical assistance given to us. To the other team members of the PMM Network in
West Africa, we say many thanks for the numerous experiences we shared during the six years of our working together. Finally, we say many thanks to our benefactors, the local Red Cross blood donors’ club in Bo Town and many other people who willingly donated their blood to save many lives. References w1x Central Statistics Office. Provisional results: 1985 census. Freetown, Sierra Leone: Central Statistics Office, 1987. w2x UNICEF. State of the world’s children, 1995. New York: Oxford University Press, 1995. w3x World Health OrganizationrUNICEF. Revised 1990 estimates of maternal mortality: a new approach by WHO a n d U N I C E F . W H O r F R H r M S M r 9 6 .1 1 ; UNICEFrPLNr96.1, April 1996. w4x BorPujehun Primary Health Care Programme Monitoring and Evaluation Unit. Prospective infant and maternal mortality survey results. Sierra Leone, 1990 Žunpublished.. w5x Thuray H, Samai O, Fofana P, Sengeh P. Establishing a cost recovery system for drugs, Bo, Sierra Leone. Int J Gynecol Obstet 1997;59ŽSuppl. 2.:141]147. w6x Fofana P, Samai O, Kebbie A, Sengeh P. Promoting use of obstetric services through community loan funds, Bo, Sierra Leone. Int J Gynecol Obstet 1997;59ŽSuppl. 2.:225]230. w7x Samai O, Sengeh P. Facilitating emergency obstetric care through transportation and communication, Bo, Sierra Leone. Int J Gynecol Obstet 1997;59ŽSuppl. 2.:157]164.