Volume 64, Issue 10 p. 1051-1060
Free Access

Global oximetry: an international anaesthesia quality improvement project

I. A. Walker

I. A. Walker

Consultant Anaesthetist, Great Ormond Street Hospital NHS Trust, London, UK

Search for more papers by this author
A. F. Merry

A. F. Merry

Professor of Anaesthesia, University of Auckland and Consultant Anaesthetist Auckland City Hospital, New Zealand

Search for more papers by this author
I. H. Wilson

I. H. Wilson

Consultant Anaesthetist, Royal Devon and Exeter NHS Foundation Trust, Exeter, UK

Search for more papers by this author
G. A. McHugh

G. A. McHugh

Senior Research Fellow, School of Nursing, Midwifery and Social Work, University of Manchester, UK

Search for more papers by this author
E. O’Sullivan

E. O’Sullivan

Consultant Anaesthetist, St James’ Hospital, Dublin, Ireland

Search for more papers by this author
G. M. Thoms

G. M. Thoms

Consultant Anaesthetist, Central Manchester NHS Foundation Trust, Manchester, UK

Search for more papers by this author
F. Nuevo

F. Nuevo

Consultant Anaesthetist, University of Santo Tomas Hospital, Philippines

Search for more papers by this author
D. K. Whitaker

D. K. Whitaker

Consultant Anaesthetist, Central Manchester NHS Foundation Trust, Manchester, UK

Search for more papers by this author
on behalf of the GO Project teams

on behalf of the GO Project teams

Results presented in part at the World Congress of Anaesthesia, Cape Town 2008.

Search for more papers by this author
First published: 02 September 2009
Citations: 53
Dr Isabeau Walker
[email protected]


Pulse oximetry is mandatory during anaesthesia in many countries, a standard endorsed by the World Health Organization ‘Safe Surgery Saves Lives’ initiative. The Association of Anaesthetists of Great Britain and Ireland, the World Federation of Societies of Anaesthesiologists and GE Healthcare collaborated in a quality improvement project over a 15-month period to investigate pulse oximetry in four pilot sites in Uganda, Vietnam, India and the Philippines, using 84 donated pulse oximeters. A substantial gap in oximeter provision was demonstrated at the start of the project. Formal training was essential for oximeter-naïve practitioners. After introduction of oximeters, logbook data were collected from over 8000 anaesthetics, and responses to desaturation were judged appropriate. Anaesthesia providers believed pulse oximeters were essential for patient safety and defined characteristics of the ideal oximeter for their setting. Robust systems for supply and maintenance of low-cost oximeters are required for sustained uptake of pulse oximetry in low- and middle-income countries.

The World Health Organization ‘Safe Surgery Saves Lives’ initiative (WHO SSSL) has identified safe surgery with safe anaesthesia as a public health priority, particularly in low- and middle-income countries [1, 2]. The mortality associated with surgery in developed countries is typically 0.4–0.8%, but may be up to 10 times higher in developing countries [1]. Anaesthesia complications are an important contributing factor to perioperative mortality, especially where resources are limited. Recent reports suggest that mortality attributable to anaesthesia in some parts of sub-Saharan Africa may be 200–1000-fold higher than in industrialised countries [3, 4].

Anaesthesia safety has advanced in recent decades in parallel with improvements in equipment and training and the introduction of guidelines for monitoring. Use of a pulse oximeter during anaesthesia is considered a mandatory monitor of oxygenation by professional anaesthesia organisations in many countries, and has recently been endorsed by the WHO and the World Federation of Societies of Anaesthesiologists (WFSA) 2008 International Standards for a Safe Practice of Anaesthesia [5, 6]. The WHO Surgical Safety Checklist, a product of the SSSL initiative, requires confirmation of the presence of a pulse oximeter during the ‘Sign-in’ prior to surgery [2]. Few anaesthetists in developed countries would consider giving an anaesthetic without a pulse oximeter, even in an emergency, yet in many developing countries anaesthesia providers routinely work without this essential monitor [7, 8].

In 2004, the Safety and Quality of Practice Committee of the WFSA identified improved access to pulse oximetry during anaesthesia as a priority for patient safety in low- and middle-income countries. The Global Oximetry (GO) Project was formed to investigate the feasibility and sustainability of pulse oximetry as a standard of care for all patients [9]. Four pilot sites were chosen where it was known that the anaesthesia providers did not have routine access to pulse oximeters. The pilot sites were in two low-income countries (Uganda and Vietnam), and in two lower middle-income countries (India, and the Philippines) [10].

The objectives of the pilot evaluation were to identify the potential oximetry ‘gap’ in the different settings; to identify the education and training needs of the different anaesthesia providers; to introduce logbooks to record the use of oximeters and to establish that providers respond appropriately to readings; and to assess the performance of two models of pulse oximeter over a 12-month period. The views of anaesthesia providers on the ideal design features of a pulse oximeter for difficult environments were obtained, and also their attitudes to pulse oximetry with respect to patient safety.

Study setting

The four pilot site projects were in selected hospitals in Uganda (population 31M), in Binh Dinh Province in Vietnam (population 2M), in Haryana State in India (population 23M) and Cebu Province in the Philippines (population 2.5M). The projects differed in emphasis and detail, reflecting local conditions, differences in anaesthesia provision, and varying opportunities for data collection in difficult environments and with limited research funding.

In Uganda, non-physician anaesthetic clinical officers (ACOs) who have received 1–2 years anaesthesia training are the main providers of anaesthesia care. Many have not used a pulse oximeter or received training in its use [7]. Pulse oximetry is not a mandated standard of monitoring in Uganda. In Binh Dinh Province Vietnam, anaesthesia is provided by four levels of practitioner: a small number of medically qualified specialists with 3 years postgraduate training in anaesthesia including pulse oximetry; physicians with 1 year of speciality training; technicians with no medical or nursing qualifications and 3 years basic training in anaesthesia; technicians with basic anaesthesia training and a further 4 years advanced anaesthesia training. Prior to the GO project, the use of pulse oximetry was not mandated in Binh Dinh Province, and was limited to a small number of donated units. In Haryana State India, provision of anaesthesia is by doctors with specialist training in anaesthesia, occasionally supported by non-specialist doctors and nurses. In 1999, the Indian Society of Anaesthesiologists recommended pulse oximetry as essential for complex surgery, and as a desired standard for other surgery, depending on the hospital facilities and the surgery performed [11]. In the Philippines, anaesthesia is administered by trained physicians only. The Philippine Society of Anaesthesiologists mandated the use of pulse oximetry during anaesthesia in 1999 and knowledge of oximetry is an essential requirement for accreditation of anaesthesia training programmes.


Relevant local approval was obtained from the Research and Ethics Committee of the Faculty of Medicine, University of Makerere, Uganda, the Uganda Ministry of Health, the Binh Dinh Department of Health, the Commissioner and Deputy Director of Health Services Haryana and the Cebu Vicente Sotto Memorial Medical Centre Investigators Review Board. The National Anaesthesia Societies of Uganda, Vietnam, India and the Philippines endorsed the project. In Uganda, individual anaesthesia providers signed informed consent to participate; in Vietnam, a memorandum of understanding was developed with the local Department of Health to support oximetry as a standard of care throughout Binh Dinh province (see below).

The GO project team members conducted site visits at the start of the project, interim visits at 3–6 months, and a final visit at 12–15 months. Information was obtained on the oximetry gap in all sites, either by interview of individual anaesthesia providers during site visits, or by hospital visits supplemented by telephone contact of lead anaesthesia providers. A pulse oximeter was expected to be present for each anaesthesia location (i.e. one per operating table), or acute care location (i.e. one per operating table, post anaesthesia care unit (PACU) bed, high dependency unit (HDU) or intensive care unit (ICU) bed). The pulse oximeter gap was defined as the difference between observed and expected numbers of pulse oximeters in the operating room (OR oximetry gap) or acute care locations (acute care oximetry gap).

A training manual was prepared to describe the theory, clinical use and maintenance of GE pulse oximeters [12]. GO project team leaders delivered the educational material during two or three half-day, face-to-face teaching sessions. Clinical scenarios were used to emphasise the action to be taken in the event of abnormal readings. The manual included multiple-choice questions (MCQs) and short answer questions (SAQs) to be completed at the start and completion of training, and at later time periods of the study.

The GE TruSat™ bedside pulse oximeter (GE Healthcare, Buckinghamshire, UK) is a lightweight oximeter with rubber bumpers to protect the screen and controls if the oximeter falls, and a backlit LCD display that can be viewed under different lighting conditions. It is simple to operate, has a rechargeable battery with back-up mains supply, and can detect low perfusion and compensate for motion artefact. TruSat™ oximeters were introduced into selected operating theatres in Uganda (20 units), Vietnam (30 units) and India (12 units). The GE TuffSat™ pulse oximeters are lightweight hand-held monitors with a rubber grip, back-lit LCD display and can detect low perfusion. They do not have rechargeable batteries, nor a mains supply. They were selected for the PACU at the Cebu Vicente Sotto Memorial Medical Centre in the Philippines (22 units) for their robust design and portability. All oximeters were provided with GE OxyTip® finger probes and cables, with some additional ear probes in Vietnam.

An anaesthesia logbook was constructed and piloted. Anaesthesia practitioners were asked to use the logbook for every case for 12 months and to record the type of surgery, the type of anaesthesia, desaturation events ≤90%, the reason for the desaturation (including artefact), and the response to and outcome of each desaturation.

Questionnaires on performance and characteristics of the ideal oximeter were developed. To construct these questionnaires, a moderated focus group was conducted at the 2006 Uganda Society of Anaesthesia meeting using a co-moderator to take notes. Eight experienced anaesthetic providers from regional and mission hospitals from around Uganda attended the focus group. A list of features describing the design of the ideal pulse oximeter for use in the local environment and the value of pulse oximetry for patient safety was generated from the discussions. A second questionnaire was designed to describe the performance of the GE TruSat™ pulse oximeter with reference to these ideal characteristics.

Details of the individual pilot sites are detailed below. Operating room gap data and performance data for the donated oximeters were obtained from all sites; the acute care gap from Uganda, Vietnam and the Philippines; education and training needs and logbook data were obtained from Uganda and Vietnam; characteristics of the ideal pulse oximeter from Uganda, Vietnam and India. Focus group discussions concerning attitudes to pulse oximetry were held in the Philippines and views on the use of pulse oximeters were obtained from Uganda.

Uganda GO pilot

The Uganda GO pilot was conducted in hospitals selected by the Uganda Ministry of Health, chosen to represent a wide geographical area and hospital type within the country (tertiary, secondary and mission hospital). The Ministry of Health selected non-physician anaesthetic officers (ACOs) who were sponsored by the GO project team to attend the Uganda Society of Anaesthesia Annual refresher course in Kampala in May 2007. They received two half-days instruction and practical demonstration of the oximeters by the Uganda GO project team. They completed MCQs/SAQs at the start of the course. They completed a second set of MCQs/SAQs and a practical demonstration of the oximeter at the end of the training; those who completed the assessments satisfactorily were issued with a certificate and enrolled into the project. The candidates were instructed in the use of the logbooks and took personal charge of the donated GE TruSat™ oximeter in August 2007. A local project officer conducted site visits at approximately 3 and 6 months. A local project officer and one of the UK investigators visited the project sites at approximately 12 months in August 2008 when the knowledge assessments and practical demonstration of the oximeters were repeated, the logbooks were collected, and any problems with the oximeter or probe were identified. The questionnaires concerning the design of the ideal pulse oximeter, the value of pulse oximetry for patient safety and the performance of the GE TruSat pulse oximeter were completed at the final site visit.

Vietnam GO pilot

The Vietnam GO pilot was conducted in Binh Dinh Province. A memorandum of understanding (MOU) was drafted between the Binh Dinh Provincial Department of Health (DoH) and the Binh Dinh GO project team in April 2007. This stated that the Vietnam GO project team would provide oximeters to all 15 hospitals in the province with educational support and maintenance for a period of 1 year, that oximeters would be used for every anaesthetic in all participating hospitals, that logbooks would be completed for each oximeter, and that the DoH would continue to provide logistical support for the use of oximetry after completion of the project as well as encouraging local hospitals to maintain and replace the oximeters as necessary. The DoH appointed an overall provincial coordinator and individual hospital coordinators in April 2007.

Training was provided at the preliminary visit by the Vietnam GO project team to 35 anaesthetic providers with the aid of interpreters during three one-day workshops in the capital city Quy Nhon, and opportunistically during hospital visits in May 2007. The educational material was translated into Vietnamese and one advanced workshop was held with the aid of the web-based, on-line simulation from the Virtual Anesthesia Machine website (http://www.vam.anesth.ufl.edu).

At the initial visit, 22 GE TruSat™ oximeters were supplied to the DoH for distribution to all 15 hospitals in the Province. The Vietnam project team visited participating hospitals during the course of the project, and again during a 2-week visit 6 months after the preliminary visit, to reinforce training, inspect logbooks, inspect the oximeters, and hold discussion groups to consider the design of oximeters and gather opinions from trainers regarding educational strategies. The project team visited all hospitals in August 2008, 15 months after the start of the project, to provide additional teaching and support for anaesthesia providers, collect logbooks and to distribute a further eight oximeters to cover the residual gap. The MOU was finalised and signed.

India GO pilot

The India GO pilot was conducted in Haryana State. In July 2007, the India GO project team and a senior anaesthetist acting as the local coordinator conducted site visits to identify suitable hospitals in the State sector for the pilot work. Interviews with key stakeholders (anaesthetists, surgeons, theatre staff and managers) were held in order to gather information on the oximetry gap in the operating room, barriers to oximeter usage and training needs. Following this, four hospitals were identified to receive the 12 donated TruSat™ oximeters.

In December 2007, the project team conducted a state-wide survey of lead anaesthetists in all 50 government hospitals in Haryana State, through hospital visits supplemented by telephone contact.

The lead anaesthetist in the four hospitals completed a questionnaire in October 2008 after 10 months experience with the 12 donated oximeters. Information was obtained about the performance of oximeters and the ideal characteristics of a pulse oximeter.

The Philippines GO pilot

The Philippines GO pilot identified the oximetry gap by national survey of anaesthetists, other physicians, nurses and nursing students through the Philippine Society of Anaesthesiologists (PSA), a telephone survey of lead anaesthetists from the hospitals in Cebu Province, also by observation by the local project team lead in the PACU, the emergency room, labour ward and ICU of the Vicente Sotto Memorial Medical Centre (VSMMC), a tertiary hospital in Cebu Province.

Three moderated focus group discussions were held at the VSMMC and a secondary hospital (the Minglanilla District Hospital). Participants included 26 nurses and 14 physicians (non-anaesthetic) from acute care areas including the emergency department, operating theatre, obstetric unit and neurosurgical and paediatric wards. Themed discussions were conducted over 1 h.

Twenty-two GE TuffSat™ oximeters were donated for use in the PACU at the VSMMC. The local project lead held a one-day training course for doctors and nurses at the VSMMC in June 2007. Questionnaires were completed by doctors and nurses prior to the introduction of the oximeters and 4 months later to gather views on the use of pulse oximeters. The donated pulse oximeters were inspected at this time and after one year.

Data analysis

Data from the questionnaires were entered and analysed using Microsoft® Excel (Microsoft Corporation, Redmond, WA, USA) and SPSS™ (SPSS Inc. Chicago, IL, USA). Descriptive and inferential statistics were used. Parametric (paired t-test) statistics were used to compare examination results for the Uganda participants at the start and at completion of the study, with the level of statistical significance at 5%. Qualitative data from focus groups and open-ended questions were theme-analysed.


In Uganda, 17 ACOs were recruited to the study from 13 hospitals. Two ACOs from one hospital failed to collect any logbook data and so were excluded from the study. Their pulse oximeters were donated to the anaesthetic department at that hospital. Data were therefore obtained from 15 ACOs working in 12 hospitals (six regional referral hospitals, two district hospitals, one military and three mission hospitals, local population served unknown). One ACO left her post so donated her oximeter to a colleague. One ACO was involved in a road traffic accident so only completed 8 months logbook data collection.

In Binh Dinh, data were obtained from all 15 hospitals in the Province. In Haryana, detail of the oximetry gap was obtained from the 39 hospitals that performed surgery. Oximeters were donated to four hospitals where an oximetry gap was identified and information on performance of oximeters was gained from these sites. Three were district hospitals with two to three operating rooms (receiving two oximeters per hospital), and one was a teaching hospital with 16 operating rooms that received six oximeters. Oximetry gap data were obtained from all 16 hospitals in Cebu Province in the Philippines and through national survey.

The oximetry gap

In the hospitals studied, the oximetry gap at the start of the project was 64%, 76%, 65% and 0% in the operating rooms in Uganda, Binh Dinh, Haryana and Cebu respectively, and ranged from 83–14% in acute care locations (including the recovery room and HDU/ICU, see Table 1). At the start of the project, a number of hospitals undertaking surgery did not have an oximeter in any location (Uganda, four of 12 hospitals; Binh Dinh, six of 15 hospitals; Haryana State, 12 of 39 hospitals). There were 22 additional oximeters that were present but not working (Uganda two, Binh Dinh two, Haryana 18), usually because of problems with the oximeter probe. Anaesthesia providers commented on that oximeters were also required in the emergency room, labour ward, surgical or paediatric wards and at times the oximeters listed above were used in these locations.

Table 1. Pulse oximetry gap at the start of the GO project (May 2007).
Uganda* (not limited by region) Vietnam (Binh Dinh Province) India (Haryana State) The Philippines (Cebu Province)
No. hospitals included in the study that provide surgery 12 15 39 16
No. operating tables 36 38 149 70
No. PACU beds 38 16 Not recorded 55
No. HDU/ICU beds 4 4 Not recorded 6
No. hospitals without access to any oximeter(% hospitals performing surgery) 4 (33) 6 (40) 12 (31) 0
No. functioning oximeters in the OR 13 9 52 70
OR oximeter gap (% anaesthetising locations) 23 (64) 29 (76) 97 (65) 0
No. functioning oximeters in all acute care locations 13 9 Not recorded 113
Acute care oximeter gap (OR, PACU, HDU, ICU) (%total locations) 65 (83) 45 (83) Not recorded 18 (14)
  • PACU, post anaesthesia care unit; HDU, high dependency unit; ICU, intensive care unit; OR, operating room.
  • *Uganda pilot carried out in different sites chosen by the Ministry of Health to represent a cross-section of the country, not limited to any specific region.

In the VSMMC, the tertiary referral hospital in Cebu Province, oximeters were available for all the operating rooms, but for only three of the 37 beds in the emergency room, two of six beds in the ICU, three of seven beds in the PACU and there were no oximeters on the labour ward. There were 296 responses from 350 questionnaires (85% response rate) sent out during the National Society of Anaesthesiologists survey in the Philippines (including 139 from anaesthesiologists). In the national survey, 275 of 296 (93%) respondents had seen a pulse oximeter (127 of the 139 (91%) anaesthesiologists), but some commented that the oximeters were their own personal possession and they were not always provided by the hospital. There were 45 of 192 respondents to the national survey (23%) who indicated that there was no oximeter in their minor operating rooms. Respondents to the national survey reported that there was no oximeter in the emergency room in 59 of 196 cases (30%), and no oximeter in labour ward in 69 of 177 cases (39%).

Education and training needs

The educational backgrounds and anaesthesia facilities for the anaesthesia providers varied greatly between study sites and are detailed in Table 2. In Uganda, drawover anaesthesia with an oxygen concentrator was used by 14 of the 15 providers, and was used exclusively by eight. Continuous flow anaesthesia using cylinders or piped oxygen supplies was used in all the other project study sites.

Table 2. Description of anaesthesia provision in pilot study sites.
Uganda* (not limited by region) Vietnam (Binh Dinh Province) India (Haryana State) The Philippines (Cebu Province)
No. hospitals included in the study that provide surgery 12 15 39 16
Regional population; million Not relevant 2 23 2.5
No. non-physician anaesthetists included in the study 15 54 0 0
Anaesthesia training for non-physician anaesthetists; years 2–3 3–7 n/a n/a
No. physician anaesthetists in these hospitals 5† 13 59 67
Anaesthesia training for physician anaesthetists; years 3 1–3 3 1–3
Pulse oximetry mandated as standard by National Anaesthesia Society No No Yes Yes
No. anaesthesia providers ‘pulse oximeter naïve’ at the start of the project 6/15 non-physicians Not known 0 0
  • *Uganda pilot carried out in different sites chosen by the Ministry of Health to represent a cross section of country, not limited to any specific region.
  • †Only non-physician anaesthetists took part in Ugandan pilot.

A number of anaesthesia providers in Uganda and Vietnam never used a pulse oximeter in their clinical practice (four of 12 hospitals in Uganda, six of 15 hospitals in Binh Dinh). All anaesthesia providers were trained in the use of pulse oximetry in the study sites in India and the Philippines.

In Uganda, the baseline knowledge of oximetry demonstrated by the SAQs completed by the ACO prior to training was poor (mean score 41%). All except one ACO completed the formal training programme at the annual refresher course (see below). Test scores improved at the end of the project for all but two providers; using a paired t-test, the average difference is 5.6 more correct answers at the end of the project (95% CI 3.8–7.5, p < 0.0005; Table 3). One of the ACOs who showed no improvement had taken over from her colleague and had not attended the face-to-face training session, although she did have access to the training materials. The ACOs in Uganda rated the training material as appropriate in all cases and commented that the scenario-based teaching was helpful. All the ACOs were able to demonstrate the basic use of the oximeter at the end of the project.

Table 3. Results of SAQs completed by anaesthesia providers at start and finish of project in Uganda.
Anaesthesia provider May 2007 Correct answers out of 18 (%) August 2008 Correct answers out of 18 (%)*
1 N/A 8 (44)†
2 3 (17) 11 (61)
3 11 (61) 14 (78)
4 6 (33) 15 (83)
5 2 (11) 13 (72)
6 4 (22) 12 (67)
7 9 (50) 14 (78)
8 9 (50) 14 (78)
9 10 (56) 9 (50)
10 6 (33) 10 (56)
11 10 (56) 15 (83)
12 9 (50) 13 (72)
13 12 (67) 14 (78)
14 6 (33) 15 (83)
15 6 (33) 13 (72)
Mean (SD) score 7.4 (3.1) 13.0 (1.9)
  • *p < 0.0005, 95% CI (3.8–7.5).
  • †Excluded from analysis as did not attend the training course.

Two thirds of anaesthesia providers (46 of 67) completed the formal training course in Binh Dinh including six of 13 doctors; the main reason for non-attendance was pressure of clinical work and geographical distance from the training centre. More focused, one-to-one training was provided to an additional 20 providers during hospital visits. Language barriers provided a substantial challenge to teaching for the visiting team, and an interpreter was required for all sessions. Scenario–based teaching and teaching using graphical, non-language based material was preferred to didactic lectures. The web-based interactive simulation was well received, particularly by those with substantial prior training. Of the 38 anaesthesia providers who were interviewed at the 15-month follow up visit, all were able to successfully demonstrate basic use of the oximeter. However, on occasion, donated oximeters were deemed broken when the problem was a completely flat battery, and the fact that this was not understood could be classified as a failure of education.

Logbook data

Logbook data were collected for over 8000 patients monitored with the donated TruSat oximeters in Uganda and Vietnam. The logbooks from Uganda relate to longitudinal data from individual anaesthetic officers for a total of 6110 cases, 4013 of which were general anaesthetics and 1867 spinal anaesthetics. There were 5490 operations classified as ASA grade 1or 2, and 32 ASA grade 4 or 5, and there were 2612 Caesarean sections. Desaturation was recorded in 427 of 3640 cases (incidence 12% cases, excluding data from 4 ACO, see below). Responses to hypoxia were appropriate when recorded in all cases (see Table 4). Four ACOs did not record any hypoxia during the study – this is not likely to have been the case, and illustrates the difficulty in collecting this type of data as a measure of performance.

Table 4. Responses to desaturation. Examples of response to hypoxia reported in logbooks from Uganda.
Clinical scenario Minimum SpO2 Response to desaturation
34 years. ASA III. Burr holes. GA and TT. 87% Patient had shallow breathing, so I did IPPV throughout operation
23 years. ASA I. Drainage of abscess. Thiopentone, halothane, oxygen. Airway obstruction (laryngospasm). 77% Did ABC and gave oxygen
20 years. ASA I. Ketamine for drainage of abscess. 88% Placed in left lateral position with jaw thrust. Good outcome
53 years. ASA I. Total abdominal hysterectomy. Spinal and pethidine. Low saturation because of sedation effect. 88% Administered oxygen
22 years. ASA II. Splenectomy. Ketamine and oxygen. 18% Attempted to intubate. Very difficult; failed intubation. Face mask and 100% oxygen with bagging. Improved with a good outcome.
4 month. ASA II. Cleft lip & palate repair. Halothane, oxygen, suxamethonium, tracheal intubation. 72% (in recovery) In recovery the patient got apnoea attacks. Naso-pharyngeal tube inserted and (airway) maintained with oxygen via nasal specs.
7 years ASA II. Excision neurofibroma of face. Oxygen, halothane, tracheal intubation. 87% Tube slipped during surgery and went into right main bronchus. Re-adjusted and ventilated.
28 years. ASA IV. Amputation and surgical toilet post road traffic accident. Severe shock & pallor. Oxygen, propofol, ether, assisted ventilation. 21% then arrested on table Gave adrenaline, cardiac massage & blood. Recovered with good outcome.
  • SpO2, oxygen saturation; ASA, American Society of Anesthesiologists (classification); GA, general anaesthesia; TT, tracheal tube ABC, Airway breathing circulation.

Logbook data were collected from 2275 anaesthetics in Binh Dinh, relating to the use of the individual oximeters rather than the activity of anaesthetists. Of these, 1981 were scheduled cases and 294 were emergencies; 2270 patients were classified as ASA 1-2 and 5 as ASA 4 or 5. Desaturation was seen in 368 cases (16% incidence); the response to desaturation was appropriate in all cases. In many cases, the problems were technical (probe not working, movement artefact, patient too cold).

Performance of the oximeters

The performance of the oximeters was assessed after 9–15 months use in Uganda, Binh Dinh, Haryana State and Cebu. Of the donated TruSat™ oximeters assessed, 36 of the 49 (73%) were in excellent working condition. One oximeter had never functioned for reasons that were unclear and were not solved by the local anaesthetists. The problems in another seven related to the battery life of the oximeters (flat, or slow to charge). Of the donated OxyTip® saturation probes, 47 of 58 (81%) were functional after 9–15 months use, and 17 of 18 ear probes were functional after 15 months. All the TuffSat™ oximeters and probes were in excellent working order at the 1-year assessment. It was noted, however, that the AA batteries used in the TuffSat™ monitors were expensive to replace, and a disadvantage in comparison to rechargeable batteries.

Characteristics of the ideal oximeter and attitudes to pulse oximetry

During the final assessment visit to each study site, both medical and non-medical anaesthesia providers in Uganda, Binh Dinh and Haryana took part in group discussions or completed questionnaires relating to the characteristics of the ideal pulse oximeter. Of 27 providers that completed questionnaires, all felt that it was important or essential that the oximeter has a long battery life, a pulse rate display, was easy clean and maintain, had an audible pulse tone that changes with saturation, was easy to use, durable, and should work well when the patient is cold and when the blood pressure is low. Flashing alarms and ability to withstand patient movement were felt to be less important. All felt that it was important or essential to use a pulse oximeter during surgery and for the safety of their patients. From the comments that were theme-analysed, the key issues which emerged were that: the oximeter gave an early warning of when things were going wrong, that it reduced the stress of administering anaesthesia to the patient, and would be beneficial to patient care on the wards, in the emergency room and labour ward. An oximeter probe that could be used for patients of all ages was desirable.

The attitudes of the anaesthetic officers in Uganda to the use of a pulse oximeter for patient safety are shown in Table 5. Focus groups on attitudes to pulse oximetry were conducted in Cebu and key themes are shown in Table 6.

Table 5. Attitudes of the anaesthetic officers in Uganda to the use of a pulse oximeter.
‘(It) really reduces stress, especially in a resource-poor setting with few staff’
‘It improves the safety of the patient by giving early warnings to the anaesthetists’
‘(I) recommend [TruSat oximeter] strongly. We operate on several cases simultaneously and we only have one pulse oximeter’
‘You panic now without it. It has taught me that patients desaturate easier than I otherwise expected’
‘(I) can’t do without it now. It’s improved my anaesthetic techniques, helped identify problems and results with a better outcome for the patient’
‘(It has) changed practice by “Early intervention to save lives”‘
‘I’m now more vigilant. Made me realise preoxygenation is paramount’
‘It’s easy to use and to teach colleagues how to use it’
‘(I have become) dependent on pulse oximetry’
‘It has simplified the practice and made it less exhausting. It helps to improve the conditions of the patient on the operating table as well as those in emergency situations’
‘You can very quickly diagnose hypoxia and do something about it. It indicates blood loss...and you can tell the surgeon’
‘I’m now more alert and more safe. I can act on emergencies very fast.’
Table 6. Attitudes to oximetry in acute care areas outside operating room – focus group discussions from the Philippines (26 nurses, 14 physicians).
Theme Summary of discussion
Need and demand for oximetry Oximeters are not freely available but are required in the ER, delivery room, for perioperative care on the wards and in the ICU, especially for trauma and paediatric patients.
Oximeters are requested by the doctors and the nurses have to share them between patients.
There are many non-functional units and wear and tear is a problem.
Physicians purchase their own oximeters for use on the wards
Workplace characteristics The supply of monitoring equipment is inadequate
The administration does not see purchase of oximeters as a priority
Maintenance of equipment is often a problem
The electrical supply is not a problem but a battery operated unit is useful for portability
Getting oximetry in place Facilities for maintenance are essential or the initiative will fail
Personnel need to be trained in use and maintenance of equipment
The hospital administration must support the introduction of oximetry
Maintaining the use of oximetry as a standard of care The clinical benefits of oximetry will be obvious and encourage the change in practice
Cost is an issue
Maintenance must be provided
Theft may be a problem for portable units
  • ER, emergency room; ICU, intensive care.

Agreements for continuing with pulse oximetry

The memorandum of understanding signed by the Department of Health in Binh Dinh Province established pulse oximetry as a new standard of care by the Department of Health for all anaesthetics administered in the Province, and recommends that the oximeters should be used, maintained and replaced as required in every hospital. At the 15-month visit, three new oximeters had been purchased, as evidence of this commitment.

Pulse oximeters donated in Uganda, Haryana and Cebu continued to be used throughout the project, demonstrating a change in practice. However, there was no formal provision for on-going maintenance or replacement. The importance of effective systems for purchasing was illustrated during discussions at site visits with medical directors and lead anaesthetists in the four hospitals studied in Haryana. Responsibility for equipment was split between medical directors and the commissioning authority at the State Capital which is many hours drive away. Individuals in these organisations may not work closely together, nor be known to each other. Faulty equipment could be out of use for very long periods. It was reported that those involved in centralised purchasing did not normally consult with end-users; nor did they necessarily include annual maintenance contracts, induction training or a re-supply system for probes when purchasing equipment. Supply was often through re-sellers of imported equipment who were later hard to trace when repairs were required. These factors were felt to be an important barrier to sustainable oximeter use. Of note, some hospitals had found that a system for careful handling of oximeter probes when not connected to the machine increased their life expectancy considerably.


We have demonstrated a substantial gap in pulse oximetry in the operating room for patients in three out of four regions in two low and two low-middle income countries, and a sizeable oximetry gap in acute care areas in all four. Binh Dinh is a province with 2 million inhabitants in the centre of Vietnam, a country of some 80 million people. There are about 50 provinces of similar size within the country, and it seems likely that many of these would have comparable oximetry gaps. India is the world’s second most populous country, and the findings in Haryana are very likely to apply to many patients undergoing anaesthesia in this country, particularly in rural areas. The considerable oximetry gap has recently been documented on a national scale in Uganda [7], and the prevailing standard of anaesthesia in Uganda is similar to that in much of Sub-Saharan Africa. From our Cebu data, it is apparent that even in some countries in which oximetry is the norm during anaesthesia, there may be a substantial unmet need in other acute care clinical situations such as the emergency department, the labour ward and the ICU. It is hard to know how large the oximetry gap is worldwide, but these findings confirm our belief that there is a substantial global deficit in monitoring by pulse oximetry for patients undergoing surgery, despite this being a recommended international standard for anaesthesia monitoring [5, 6].

The need for education and training of practitioners was clearly demonstrated in areas where oximeters were not in routine use. We have identified features of an ideal pulse oximeter for use in settings with limited resources. We have also demonstrated that currently available oximeters can be used in difficult settings with constrained overall resources, and that with relatively little additional education, anaesthesia providers can adopt this form of monitoring as a routine part of practice and respond appropriately to the information provided. Both oximeters used in the project were highly rated; they are easy to use and robust. The use of replaceable batteries in the TuffSat™ was a limitation (because of the ongoing expense of replacing these); the inbuilt rechargeable batteries TruSat™ seemed very suitable. The probes used were more vulnerable than the oximeters, but proved more durable than expected. At the end of one year, a substantial majority of the original probes were still functioning and in use and our impression was that all of this equipment was highly valued and greatly cared for. Interview and focus group data confirmed that participating providers were uniformly positive about pulse oximetry, and strongly convinced of its value in contributing to patient safety.

Changing practice will only be of value if the change is sustained. It is not reasonable to continue donating equipment and support to a region indefinitely, and the donor model is not suitable for middle-income countries. The key to sustained adoption of oximetry as a standard of care lies in working with local providers, hospital administrators, and central government officials to ensure that pulse oximetry is recognised as a standard of care, anaesthesia providers are trained appropriately, local financial commitment is made to purchase, maintain and replace probes and oximeters as needed, and clear systems are in place to achieve this. In Binh Dinh, we have demonstrated the possibility of doing this. This model may be thought of as akin to employing a catalyst. Probably this province would have adopted oximetry in due course without our intervention; our project has considerably accelerated this process. Ongoing support in the form of visits, monitoring and trouble shooting may be of value, but the ultimate aim must be to develop the local systems for safe standards of anaesthesia care. Our project was helpful in highlighting local deficits and needs for education and training.

This study is a quality improvement initiative, but has a number of limitations. It has been run with minimal financial resources. The data are primarily descriptive, and mostly qualitative, and by nature, there is no before and after data. While some of the information has been collected in a structured fashion, much has been collected opportunistically. The participating countries were not selected randomly; they were chosen because of anecdotal knowledge that gaps were present, because of their receptivity to the project, and because of existing linkages and contacts with these areas. Nevertheless, they probably represent a reasonable cross section of the situations in which an oximetry gap is an important public health problem. The sites were sufficiently diverse to elicit a remarkable range of information about the relevant issues, and this is one of the study’s strengths.

The ideal measure for any safety intervention would be the effect on patient outcome. This is very difficult to do in relation to safety in anaesthesia, primarily because of the large number of patients and considerable resource needed. The validity of measuring structure and process to demonstrate the value of safety interventions in healthcare is well established [13]. Deploying over 80 oximeters and confirming that the majority were working and in use a year later represents a substantial change in the structure and process of these anaesthesia services. Improvement in participants’ understanding of oximetry can also be listed on the credit side of the structure account as the knowledge and skill of the human resource is of paramount importance to anaesthesia safety.

An affordable and suitable oximeter is obviously part of the solution to the global oximetry gap, but the experience reported here shows that addressing the gap in oximetry in any particular region requires more than this. Replacing probes might prove more expensive over the life of an oximeter than the initial cost of the oximeter itself. Different probes are needed for children, adults and neonates, and it would be very helpful if manufacturers were able to produce a universal probe that was also easy to use and robust. The prime importance of maintenance and ready access to servicing of oximeters and replacement of damaged probes was very clear. If equipment is to be of long term value, a local in-country agent is essential, as are clear arrangements for contacting that agent when servicing or replacement is required, and a budget to pay for this. Without these features, the equipment will probably be used until it breaks and then placed in a cupboard and forgotten, and examples of this were seen. Standardisation of equipment to one or two suppliers who can then provide strong support makes sense.

New technology should only be introduced with an educational package. Those responsible for delivering our educational package found it practical and straightforward to teach. Nevertheless, improvements could be made, particularly in relation to crossing language barriers. Those with expertise and practical experience with pulse oximetry should deliver the education, ideally in an interactive format. More work on providing the resources for this would be worthwhile. Before attempting to undertake any project of this type, a formal educational needs assessment would be very important. Our study highlighted the variation in educational need that exists between regions. In Haryana and Cebu, medically trained anaesthetists administer anaesthesia, most of whom were familiar with the principles of oximetry. In Uganda, six of 15 anaesthesia providers had never seen a pulse oximeter, while in Binh Dinh, education was particularly challenging because of language barriers. Material specifically designed for the educational needs of anaesthesia providers in different settings would be of value.

Many providers saw the need for oximetry in acute care settings as equal to that in the operating room (OR); it is interesting, therefore, that meeting the need in the OR led to greater use of oximetry in PACU and other acute care settings as providers and funders became convinced of its value.

The World Health Organization has now included universal pulse oximetry as a formal part of its Safe Surgery Saves Lives initiative. We believe that this is one of the most important programmes in anaesthesia safety in the last few decades. The data presented here endorse the need for this ambitious international project and we hope will help to enable ‘Global Oximetry’ to become a reality.

*Members of the GO project teams

A tripartite relationship was developed for the GO project between the Association of Anaesthetists of Great Britain and Ireland (AAGBI), the World Federation of Societies of Anaesthesiologists (WFSA) and General Electric Healthcare (GE). Collaborating institutions included University of Manchester (UK), University of Auckland (NZ), the New Zealand Society of Anaesthetists (NZSA), the New Zealand Vietnam Health Trust (NZVHT) and the International Clinical Epidemiology Network (INCLEN).

GE Healthcare

Leena Pesu, Mark Phillips, Colin Hughes.

The Uganda GO Pilot

Uganda: David Snell, Clare Labulwa, Margaret Okello, Sarah Hodges. UK: Isabeau Walker (AAGBI), Ellen O’Sullivan (AAGBI), Iain Wilson (AAGBI, WFSA).

The Vietnam GO Pilot

Alan Merry (WFSA, University of Auckland), David Morris (NZVHT), Lien Morris, John McDougall (NZVHT), Ha Anh Thach, Global Oximetry (GO) Binh Dinh Project Coordinator, Department of Health, Binh Dinh Province, David Chamley (NZVHT).

Steuart Henderson (NZVHT), Pat Alley (NZVHT), Nguyen Huu Tu (Hanoi), Barbara Dawson (NZVHT representative in Quy Nhon), Philippa Bascand NZSA), Anne deBres, VSA, Binh Dinh.

The India GO Pilot

Gavin Thoms (WFSA & University of Manchester), Gretl McHugh (University of Manchester), Sarla Hooda (PGMIS Rohtak – lead in Haryana), Rajiv Singh (OSA Manager GE Healthcare, India), Narendra Arora (INCLEN, Delhi, India).

The Philippines GO Pilot

Florian Nuevo, Emerson Donaldo, Glenn Sarucam, William Cesar Ortiz, Francis Jayson Villegas.


The authors are grateful for funding received from the Association of Anaesthetists of Great Britain and Ireland, the World Federation of Societies of Anaesthesthesiologists, the New Zealand Vietnam Trust, The New Zealand Society of Anaesthetists and the University of Manchester and in particular to GE Healthcare Ltd for donating all the oximeters and oximeter probes used in the project.