|Year : 2016 | Volume
| Issue : 3 | Page : 146-152
Appropriate anaesthesia technology for improved access to surgical services in resource constrained facilities in Northern Nigeria: A pilot study of the Universal Anaesthesia Machine
Alhassan D Mohammed1, Henry V Doctor2, Godwin Y Afenyadu3, Jamilu Tukur4
1 Department of Anaesthesiology and Intensive Care, Bayero University/Aminu Kano Teaching Hospital, Kano, Nigeria
2 World Health Organization, Regional Office for the Eastern Mediterranean, Cairo, Egypt
3 Department of Population and Family Health, Columbia University, New York, USA
4 Department of Obstetrics and Gynaecology, Bayero University/Aminu Kano Teaching Hospital, Kano, Nigeria
|Date of Web Publication||17-Mar-2017|
Alhassan D Mohammed
Department of Anaesthesiology and Intensive Care, Bayero University/Aminu Kano Teaching Hospital, P.M.B. 3452, Kano
Source of Support: None, Conflict of Interest: None
Background: The provision of safe anesthesia in developing countries is often challenging due to multiple health systems and infrastructural challenges. We explored the suitability and reliability of the Universal Anaesthesia Machine (UAM) in Northern Nigeria.
Patients and Methods: Over an 18-month period from 1st June 2012 to 30th November 2013, Nurse anesthetists and biomedical technicians were trained on the use, maintenance, and installation of the UAM in the study facilities. Patients requiring general anesthesia (GA) were then randomly assigned to the UAM or other forms of GA. Availability, ease of ues ans safety were monitored.
Results: A total of 1562 patients had anesthesia administered with 54.8% using UAM. Cesarean section due to obstructed labor was the most common indication. There was no malfunctioning of the UAM and its use was preferred over other forms of GA. Complications were less than with other form of GA provided by the same staff.
Conclusion: The UAM is suitable for health facilities with limited equipment and manpower. It has the potential to improve access to surgery and emergency obstetric care as it can be safely used by non-physician anesthetists.
Keywords: Emergency obstetric care, general anesthesia, maternal mortality, Universal Anaesthesia Machine
|How to cite this article:|
Mohammed AD, Doctor HV, Afenyadu GY, Tukur J. Appropriate anaesthesia technology for improved access to surgical services in resource constrained facilities in Northern Nigeria: A pilot study of the Universal Anaesthesia Machine. Arch Int Surg 2016;6:146-52
|How to cite this URL:|
Mohammed AD, Doctor HV, Afenyadu GY, Tukur J. Appropriate anaesthesia technology for improved access to surgical services in resource constrained facilities in Northern Nigeria: A pilot study of the Universal Anaesthesia Machine. Arch Int Surg [serial online] 2016 [cited 2018 Nov 18];6:146-52. Available from: http://www.archintsurg.org/text.asp?2016/6/3/146/202367
| Introduction|| |
Many developing countries, especially those in sub-Saharan Africa (SSA), face a critical shortage of healthcare professionals and resources. Health systems are stretched by human immunodeficiency virus (HIV)/acquired immunodeficiency syndrome (AIDS), malaria, loss of trained health personnel to the developed world, insecurity, and economic effects of long-term conflicts., Anesthesia services are particularly vulnerable to these pressures. They are often provided by nonphysicians and regarded as a low priority that lacks the traction to demand access to resources.
Since the 1980s, technological advances and improvements in training protocols have led to improvements in the quality and safety of anesthesia. The Safe Anesthesia Working Group of the World Health Organization's ''Safe Surgery Saves Lives'' global initiative updated the WFSA's 1992 International Standards for the Safe Practice of Anaesthesia which address, in detail, the organization, support, practices, and infrastructure for anesthesia care. Despite these improvements, rural SSA still lacks even the basic anesthetic equipment standards. Millions of people in developing countries do not have access to safe anesthesia and pain relief during surgery and childbirth.,
Some of the key goals of the 2000 Millennium Declaration were the Millennium Development Goal (MDG) 4 targeting a two-third reduction in under-5 mortality and MDG 5 targeting a 75% reduction in maternal mortality, both from 1990 figures. Despite global progress in a number of MDGs, these two MDGs are far from being reached by many countries,, such as Nigeria that was ranked 142 based on the 2010 Human Development Index. Nigeria has one of the highest maternal mortality ratios (MMR) (deaths per 100,000 live births) with ratios in the northern part exceeding 1000 compared with 300 in the southern part of the country.
More than three decades, there is the fear that much has not been achieved in preventive primary health care aimed at reduction in maternal mortality in the poorest countries. In Nigeria, results from the 2013 Demographic and Health Survey (DHS) painted a grim picture. The MMR was estimated at 576 deaths per 100,000 live births (LBs), a slight increase from the MMR of 545 deaths per 100,000 LBs in the 2008 DHS. Acceptance of modern family planning (FP) is low with 10% of married women using modern contraceptive methods, whereas the unmet need for FP was 16%. Skilled birth attendance rates were low with 41% of women in the North West zone of Nigeria having received antenatal care (ANC) from a skilled health provider whereas 11.5% were delivered by a skilled health provider.
Low FP uptake, skilled birth attendance, and lack of access to emergency obstetric care (EmoC) contribute immensely to maternal mortality. Hemorrhage, eclampsia, obstructed labor, and unsafe abortion are some of the obstetric complications that lead to death. Cesarean section (CS), hysterectomy, vacuum extraction, forceps delivery, and removal of retained placenta are some of the key interventions for most of these obstetric emergencies. These procedures require one form of anesthesia or the other that could be in the form of general, regional, or local anesthesia. The provision of safe anesthesia requires appropriate and functional equipment and skilled human resource.
In SSA, CS rates are as low as 1.5% of all births against the expected rate of 5–15%. Maternal deaths multiply several folds if the procedure is not done when required. In developed countries, CS is not only available but its safety has also improved as general anesthesia (GA) has given way to regional anesthesia. However, resuscitation followed by GA is often times the only choice for patients requiring CS in developing countries.
There is an acute shortage of anesthetists in West Africa and northern Nigeria in particular. Persistent power outages, inadequate funding, regular breakdown of equipment, and absence of spare parts contribute to poor quality anesthetic services in Nigeria. Most of the available state-of-the-art surgical or anesthetic equipment quickly become dysfunctional due to lack of preventive maintenance or unavailability of rather expensive replaceable spare parts. This situation is compounded by acute shortage of qualified biomedical engineers and technicians required to maintain and repair anesthetic equipment. A workshop on anesthesia recommended only the use of equipment designed to cope with local conditions to improve safety and quality of anesthesia. Therefore, a user-friendly anesthetic machine that is affordable, efficient, safe, and easily serviced is not only ideal for health facilities in developing countries but also capable of reducing morbidity and mortality associated with surgery or complications of childbirth.
The Universal Anesthesia Machine
The UAM was developed for use in developing countries where the supply of compressed gases and electricity are often unreliable. The components of the UAM have been safely used since 1999, and over 24000 surgical cases have been conducted over a 10-year period in Malawi. This high specification and internationally recognized machine heralds new possibilities for the provision of safe anesthesia in developing countries. It differs from the standard Boyle's machine in that it has an oxygen concentrator, draw-over vaporizer, bellows, and balloon valve. The system provides continuous flow anesthesia, reverting to draw-over mode if air is entrained or if electricity fails with the vaporizer and bellows continuing to function normally. In both modes, oxygen can alternatively be provided via cylinder, pipeline, or the side emergency outlet. All parts of the UAM are designed to have minimal to no service requirements.
Piloting the Universal Anesthesia Machine
Due to poor access to affordable, safe, and quality surgical and obstetric an aesthetic care in northern Nigeria, there was a need to systematically introduce the UAM to potential stakeholders by demonstrating its functional suitability for the Nigerian environment. Specifically, the study explored the suitability and reliability of the UAM when it was used for surgery requiring GA and then compared with other forms of available GA. The pilot study was implemented in Kano, Katsina, and Zamfara States in Northern Nigeria. These states were selected as part of the program to revitalize routine immunization and improve maternal and child health in selected states in Northern Nigeria (PRRINN-MNCH Programme). The PRRINN-MNCH Program was funded by the United Kingdom Department for International Development (UKAid) and the Government of Norway. Located in the North-Western zone of Nigeria, Kano, Katsina, and Zamfara States had a population of 9.3 million, 5.8 million, and 3.2 million, respectively, according to the 2006 population census.
| Patients and Methods|| |
Murtala Muhammed Specialist Hospital (MMSH in Kano State), General Hospital Kaura Namoda (Zamfara State), and General Hospital Funtua (Katsina State) were selected for the pilot study to reflect the urban (MMSH Kano) and rural Northern Nigeria settings (Zamfara and Katsina hospitals). A preintervention assessment of the selected facilities was conducted using the World Health Organization (WHO) Tool for Assessment of Emergency and Essential Surgical Care to collect information on the number, skills, level of training of relevant healthcare personnel, as well as availability and functionality of equipment for surgical and obstetric care. Information on surgical caseload, including CS, in 12 months prior to the pilot was also obtained.
Background characteristics of pilot health facilities
The General Hospital at Kaura Namoda is a 70-bed facility expected to serve 281,367 people (2006 census). Total admissions and outpatients in 2011 exceeded 2000 and 5000, respectively. There were 4 general duty medical officers (3 were transferred before the pilot), 2 theatre nurses (1 of them trained on the job to administer anesthesia), and 2 biomedical technicians. It had one operating room for all surgical procedures, which had two nonfunctional Boyles anesthesia machines due to lack of oxygen and maintenance. Surgical procedures ranged from 301 to 400 with CS being the most common. There was heavy reliance on ketamine for only GA. It is the poorest of the three study facilities in terms of equipment, manpower, and electricity supply.
The 200-bed General Hospital Funtua serves 225,571 people (2006 census). Outpatients and inpatients in 2011 numbered 175,987 and 13538, respectively. Surgical procedures were in excess of 2000, with CS being the most common. There were 4 medical officers, 1 visiting obstetrician and gynecologist, 3 nurse anesthetists, and 1 biomedical technician. All the surgeries were performed in one operating room. There were two Boyles anesthesia machines that were often not used due to irregular and insufficient supply of oxygen, which encouraged reliance on ketamine only GA. Electricity supply was erratic but complemented by a standby generator.
Murtala Muhammed Specialist Hospital, Kano served over 9 million people (2006 census). In 2011, a total of 27301 patients were admitted; 5.7 million were outpatients, and 4310 surgical procedures were performed. There were 2 qualified surgeons, 2 full-time obstetricians (2 part-time), and 18 medical officers. There was 1 full-time physician anesthesiologist, 1 visiting anesthesiologist, 10 nurse anesthetists, and 1 biomedical technician. Surgical procedures performed include laparotomies, CS, herniorrhaphy, and appendectomy. There were 7 operating theatres with functional anesthetic machines, which were mostly donated. Oxygen supply was adequate, but ketamine as a sole anesthetic was often used due to poor maintenance of the anesthetic machines. Electricity supply was adequate.
A 3-day training workshop was conducted for anesthetists and biomedical technicians from the three study sites regarding the installation, use, and maintenance of the UAM. This was followed by installation of the UAM in each of the facilities. A step-down 1-day workshop on data collection and postoperative complications was also conducted at each facility to standardize data collection.
All patients scheduled for elective or emergency surgical procedure that required GA and who had consented were recruited into the study. A double-blind procedure was used to allocate the patients to either UAM or other forms of GA available at the facility.
Data regarding age, sex, height, weight, and all relevant information derived during the preoperative, intraoperative, and postoperative period were collected. The preoperative data included the type of procedure, physical condition of the patient, any associated medical illnesses, indication for surgery, effectiveness and functionality of the UAM, fetal outcome, and patient outcome (morbidity and mortality). The anesthetist recorded all the required information on a data sheet.
Regular monitoring and evaluation visits were conducted by a trained consultant anesthetist to all the pilot sites to ensure compliance with the study protocol, to identify and resolve logistics and other challenges, and to document any malfunction or breakdown of the equipment. Qualitative data regarding the suitability and reliability of the UAM was also obtained from the users throughout the study. Pilot data collection covered an 18-month period (1 June 2012 to 30 November 2013). The quantitative data were analyzed using STATA version 12 (OES Medical, Witney, United Kingdom).
A full implementation research protocol was developed and approvals were obtained from the Research Ethics Committees of the State Ministries of Health in each of the pilot states.
| Results|| |
Selected characteristics of patients
A total of 1562 patients were recruited of which 1164 (representing 74.5%) were females. Out of the 1164 females, 1000 females (representing 85.9%) presented with obstetric cases. Approximately 76% of the obstetric cases (i.e., 759 out of 1000) had no antenatal care prior to presentation. There was rural-urban inequity in the distribution of specialist surgeons who operated on the patients. General duty medical officers performed majority of the surgical operations. For example, medical officers performed 86.7% of the 758 cases in Katsina, 100% of 281 cases in Zamfara, and 84.5% of the 523 cases in Kano. Across all facilities, medical officers performed 86.9% of all the 1562 surgical cases. These differences were statistically significant (Chi-square = 17.90; P < 0.001).
Anesthetic technique and relative use of Universal Anesthesia Machine
Approximately 75% of all patients (i.e., 1172 out of 1562) across all sites had inhalational GA administered via facemask (GAFM) whereas 25% (i.e., 391 out of 1562) of all the patients had either local anesthetic (LA) or total intravenous anesthesia using ketamine.
There was a reasonable split in the use of UAM (54.8%) against nonUAM (45.2%) across all facilities (N = 1562). The split was similar in each facility [Table 1], with Kaura Namoda having the lowest utilization (N = 281) compared to Funtua (N = 758) and Murtala (N = 523).
|Table 1: General anesthesia by health facility among all patients, northern Nigeria|
Click here to view
Type of obstetric surgical cases
Obstructed labor, breech presentation, antepartum hemorrhage, and eclampsia were the major obstetric cases that presented for surgery and constituted 79.7% (i.e., 797 out of 1000 cases) and by their nature were emergencies [Table 2]. Obstructed labor alone accounted for slightly more than half (53.9%) of the obstetric cases (i.e. 539 out 1000). The least common presentations were premature rupture of the fetal membrane and retained placenta (0.1%, n = 1). The pattern of presentation was similar by facility.
|Table 2: Percent of obstetric cases presented for surgery by facility, northern Nigeria|
Click here to view
Parity of obstetric patients
Across all facilities, 99.6% of all (N = 232) CS patients were multipara, with a mean parity (children ever born) of 3.7. The highest mean parity was observed in Zamfara (6.8; N = 4) followed by Kano (3.8; N = 208) and Katsina (2.8; N = 22). These differences were statistically significant (chi square = 15.29; P = 0.018).
Preoperative physical status
Approximately two-thirds (65.6%) of the emergency cases had no concurrent medical illness (American Society of Anesthesiologists (ASA) 1E); mild associated systemic illness, ASA 2E, constituted 20.5%. Approximately 10% were poor whereas those with very poor physical condition constituted 3.8% [Figure 1]. The differences between physical status and facility were statistically significant (chi square = 187.27; P < 0.001).
|Figure 1: American society of anesthesiologists (ASA) classification of preoperative physical status|
Click here to view
Duration of anesthetic exposure
The mean duration of anesthetic exposure across all facilities was under one hour (UAM = 54.5 minutes; NonUAM = 50.8 minutes; all machines = 54.9 minutes), and the differences were not statistically significant according to the type of machine. Throughout the study period, there was no equipment malfunction or failure observed from the use of the UAM.
Feedback from users
The UAM users believed that it is a unique equipment with many advantages over and above the various continuous flow anesthesia machines that rely on compressed oxygen and electricity. Users described the machine as versatile, simple, safe, and economically beneficial [Table 3]. Users from Funtua described the UAM as suiting the nature (poor resource) of their busy hospital. They observed a remarkable improvement in perioperative care because they were able to accommodate cases that would normally be referred elsewhere for inhalational anesthesia due to inadequate and unreliable supply of compressed oxygen needed to use the available Boyles machines. The heavy reliance on ketamine in their centre was not a concern among users. Owing to its simplicity and multipurpose functions, users considered UAM as the best anesthetic machine they ever worked with and urged their State Ministry of Health to acquire and install the machine in other hospitals.
At MMSH in Kano, there were similar observations but with an additional economic perspective. In the maternity theatre, approximately 4–5 giant cylinders of oxygen were consumed weekly, but with the UAM, about 60000 Nigerian Naira (USD 370) were saved weekly. Users confirmed that the UAM did not require maintenance besides routine cleaning and pre-use check. Most importantly, the UAM did not malfunction throughout the pilot period.
Data collection was challenging in the first few weeks but significantly improved during monitoring visits. Another challenge was randomization because users found the UAM user-friendly and regarded it as a better alternative to the ketamine anesthesia. Consequently, the bias toward the UAM was detected early and rectified by re-emphasizing the importance of randomization to the users. Furthermore, a prerandomization chart was developed for each site that enhanced adherence. A senior staff at each site was selected to provide daily supervision and ensure compliance. These measures were applied to the General hospital Kaura Namoda as well, which recorded the lowest utilization.
| Discussion|| |
The preintervention assessment of the pilot facilities generally mirrored the resource limitations experienced by many rural health facilities in developing countries. For example, the least resourced facility, Kaura Namoda General Hospital, had 1 nurse anesthetist trained on the UAM and 3 of the 4 medical officers, who could perform life-saving surgical procedures, were transferred out of the facility. Consequently, the GA services case load at this facility was significantly lower than expected for its catchment population. Like in many SSA countries, general duty medical officers performed surgical procedures and GA was administered by nurse anesthetists using a facemask.,,,,,,,
The inadequate surgical and anesthesia resources observed in our study are also consistent with findings elsewhere. Most SSA health facilities often have no plans and budget for sustainable procurement, replacement, repairs, and preventive maintenance of appropriate equipment. The appropriateness of anesthetic equipment is best informed by the context of erratic power and compressed gas supply, lack of highly skilled anesthetists, poor procurement practices, and few biomedical staff.
Majority of our patients with obstetric complications had inhalational anesthesia administered via a facemask. Although endotracheal tubes and other ancillary anesthetic equipment were available, GA with endotracheal intubation could not be instituted by the nurse anesthetists due to lack of or apparent loss of skills and competencies to intubate patients requiring GA. GA without endotracheal intubation could potentially increase the risk of aspiration pneumonitis and the associated increased morbidity and mortality in the preoperative period.,
Five maternal deaths were recorded (2 from UAM; 3 from nonUAM). These deaths were not significantly related to the type of the anesthetic machine used, but could be attributed to other factors. Preoperative hemoglobin or hematocrit, serum urea and electrolytes, and glucose levels, for example, may have been contributory but were not recorded. One of the limitations encountered in this study was inadequate preoperative data. Anesthetic chart was only being used at Funtua by only one nurse anesthetist. There was no evidence of its use in the other two facilities.
Risk factors of anesthesia mortality are well known. In the United Kingdom, the Confidential Enquiry into Maternal Deaths cited inexperience of the anesthetists, failed airway management, obesity, and the administration of oxytocin as contributing factors., It was also reported that the increasing proportion of patients managed under regional anesthesia, together with the reduction of hours by the trainees, could result in decreased skill levels of GA providers. In contrast, a study in a tertiary hospital in Nigeria revealed that, like in many other African countries, the predisposing factors to anesthesia mortality include failed airway management, inadequate supervision of junior anesthetists and insufficient monitoring. Our study, however, showed serious lack of human resources and the appropriate equipment for safe anesthetic practice. This is especially the case in Kaura Namoda in Zamfara State. Any systematic audit of deaths in the preoperative setting should consider the available resources (skilled manpower and equipment), the level of experience of the anesthesia provider, and the preoperative physical condition of the patient.
Physician anesthetists are scarce in developing countries. WHO estimates that Africa has a shortfall of 817,992 doctors, nurses, and midwives. Task shifting has been advocated as a means of mitigating the shortfall. In the absence of the physician anesthetists, appropriately trained nurses could provide anesthesia, especially in very poor resource settings.
This study showed that the UAM could be introduced to health facilities in resource-constrained northern Nigeria or similar settings to provide safe and reliable anesthesia because it was considered user-friendly by users, and as demonstrated in this study, health workers other than physicians could be trained on its use.
| Conclusions|| |
The UAM is suitable for health facilities in resource-constrained northern Nigeria. The preference of UAM utilization by the users, the fact that no maintenance was needed nor was any malfunction recorded, and the significant funds saved from oxygen utilization underscored the UAM's attributes of simplicity, safety, and economic benefit. The GA using face mask that was used in majority of the cases is a critical step in the transition from ketamine only based anesthesia to inhalational anesthesia facilitated by the UAM. Because a majority of the cases were obstetric, the UAM has the potential to improve access to EmoC as a direct consequence of strengthening anesthesia services through appropriate technology. If the UAM is widely used across rural northern Nigeria, attainment of MDG 5 could be accelerated. However, future studies on this topic would need to collect comprehensive data on maternal and fetal outcomes to assess the impact of the UAM on these outcomes–something that this study was not able to document.
A number of challenges were encountered during the study. First, implementation of randomization in the first few weeks was challenging with a bias in favor of UAM anesthesia. Although measures to encourage compliance of the randomization protocol were instituted with remarkable improvement during the rest of the study period, results could still be biased. Second, important basic data on maternal and fetal outcomes were grossly deficient or unavailable, and therefore could not be analyzed. Third, it was impossible to standardize the practice of anesthesia among the nurses in the three study facilities. Indeed, there were marked individual differences between nurse anesthetists in Murtala Muhammed Specialist Hospital (MMSH) and Funtua in their approach to provision of anesthesia. Lastly, the UAM was installed in a maternity theatre at MMSH and the cases were purely obstetric (including ectopic and evacuations) whereas the cases from the other two facilities included nonobstetric cases because the UAM was installed in theatres being used for all cases. Despite these limitations, the results from this study show that the use of the UAM across rural northern Nigeria can accelerate attainment of MDG 5.
The authors would like to thank the management and staff of the three pilot facilities for taking part in this project. Technical and logistical support was received from Gradian Health Systems, PRRINN-MNCH Programme, and the State Ministries of Health in Kano, Katsina, and Zamfara States, Northern Nigeria.
None of the authors was associated in the concept, production, selling or marketing of the Universal Anaesthesia Machine. However, the research was funded by Gradian Health Systems, New York, the organization that funded the production/manufacture of the UAM. All opinion expressed are, however, the authors'.
Financial support and sponsorship
Project was funded by Gradian Health Systems.
Conflicts of interest
There are no conflicts of interest.
| References|| |
Hodges SC, Mijumbi C, Okello M, McCormick BA, Walker IA, Wilson IH. Anaesthesia services in developing countries defining the problems. Anaesthesia 2007;62:4-11.
Merry AF, Cooper JB, Soyannwo O, Wilson IH, Eichhorn JH. An iterative process of global quality improvement: The International Standards for a Safe Practice of Anesthesia 2010. Can J Anesth/J Can Anesth 2010;57:1021-6
Soyannwo OA, Elegbe EO. Anaesthetic Manpower in West Africa – Which way forward? World Anaesth 2000;4:1-2.
Bhutta ZA, Black RE. Global Maternal, Newborn, and Child Health- So Near and Yet So Far. N Engl J Med 2013;369:2226-35.
Countdown to 2015 for maternal, newborn and child survival: Accountability for maternal, newborn, and child survival. Geneva: World Health Organisation; 2013.
Fenton PM. Maternal mortality and anaesthesia technology in the 21st
century. Anaesthesia News; April 2010, issue 273. p. 5-8.
Fenton PM. Maternal deaths and anaesthesia technology in the 21st
century. World Anaesth News 2010;11:12-6.
National Population Commission Nigeria and ICF International 2014. Nigeria Demographic and Health Survey 2014. Abuja, Nigeria: National Population Commission and Rockville, Maryland, USA: NPC and ICF International.
Onwurzirike BK, Uzochukwu BS. Knowledge, attitude and practice of family planning amongst women in a high density and low income urban area of Enugu, Nigeria. Afr J Reprod Health 2001;5:83-9.
Fenton PM, Whitty CJ, Reynolds F. Caesarean section in Malawi: Prospective study of early maternal and perinatal mortality. Br Med J 2003;327:587.
Murekatete C, Kapatuka K. All Africa Anaesthesia Congress, Nairobi, 2009: Equipment Workshop Report. World Anaesth 2010;11:19-22.
Van Hasselt GL, Barr KG. The first Universal Anaesthesia Machine: An evaluation. J Anaesth Pract 2011;4:128-134.
Dubowitz G, Detlefs S, McQueen KA. Global anesthesia Workforce Crisis: A preliminary Survey Revealing Shortages Contributing to Undesirable Outcomes and Unsafe Practices. World J Surg 2010;34:438-44.
Collins C, Gurung A. Anaesthesia for Caesarean Section. Update Anaesth 1998;9:7-17.
Dyer RA, Reed AR, James MF. Obstetric anaesthesia in low-resource settings. Best Prac Res Clin Obstet Gynaecol 2010;24:401-12.
Lewis G (ed). The Confidential Enquiry into Maternal and Child Health (CEMACH). Saving Mothers' Lives: Reviewing maternal deaths to make motherhood safer-2003-2005. 2007.
Lyons G. Saving mothers' lives: Confidential enquiry into maternal and child health 2003-2005. Intl J Obstet Anesth 2008;17:103-5.
Searle RD, Lyons G. Vanishing experience in training for obstetric general anaesthesia: An observational study. Intl J Obstet Anesth 2008;17:233-7.
Enohumah KO, Imarengiaye CO. Factors associated with anaesthesia-related maternal mortality in a tertiary hospital in Nigeria. Acta Anaesthesiol Scand 2006;50:206-10.
[Table 1], [Table 2], [Table 3]