Indicator Collection Methodology
| Country | Population | Bellwether hospitals (n) | Indicator 1 methodology | Indicator 2 methodology | Indicator 3 methodology | Indicator 4 methodology | |
| 2-Hour access zone | Population distribution | SAO density/100 000 | Surgical volume Cases/100 000 | POMR | |||
| Australia | 23 946 000 | 112 | 200 km radius around bellwether hospitals | Government population statistics | Specialist Medical Workforce Database Medical Board of Australia Australian Institute of Health and Welfare website | Data obtained from already established centralised government database maintained by the Australian Institute of Health and Welfare | Data obtained from Australian Institute of Health and Welfare |
| Timor Leste | 1 300 000 | 3 | Manual contour line created | Government population statistics | Direct contact with clinical directors of hospitals by members of working group | Data obtained from theatre logbook and/or anaesthetic registry of each bellwether hospital; collected by telephone or email enquiry to each hospital (no pre-existing centralised database available) | Prospective data collection in single hospital (2015), representing >85% of surgery performed in the country |
| Fiji | 933 000 | 5 | Manual contour line determined by Fiji Bureau of Statistics | Government population statistics | Medical Workforce Database (Database of the Fiji Medical Council) | Data obtained from theatre logbook registry for each bellwether hospital (5 in total); collected by phone or email enquiry to each hospital (no pre-existing centralised database available) | Retrospective review of patient records in each bellwether hospital to determine POMR within 7 days No pre-existing system to report national POMR |
| Tonga | 103 000 | 1 | Island mapping | Government population statistics | Direct contact and personal knowledge of members of collaborative working group | Pre-existing centralised database maintained by Ministry of Health Established reporting system with data collected from single hospital theatre registry | Data available from Ministry of Health Established reporting system with data collected from single hospital theatre registry and hospital mortality audit/registry |
| Samoa | 187 000 | 1 | Island mapping | Government population statistics | Direct contact and personal knowledge of members of collaborative working group | Data obtained from theatre logbook and/or anaesthetic registry of single bellwether hospital (no pre-existing centralised database available) | Mortality data obtained from hospital death certificates No pre-existing process for collecting POMR |
| Vanuatu | 260 815 | 2 | Manual contour line for each of 3 bellwether hospitals and island methodology | Government population statistics | Direct contact and personal knowledge of members of collaborative working group | Theatre and anaesthetic registers in each of the bellwether hospitals (3-year period 2012–2015 used to calculate annual case volume) (no pre-existing centralised database available) Central health information system exists but considered unreliable for these purposes due to inaccurate and incomplete data | Individual case volumes and postsurgery deaths collected for each bellwether hospital using theatre and anaesthetic registry of procedures checked against hospital records of inpatient deaths and correlated with death certificates National death registry available but of limited use because no record of surgical procedure on this registry |
| Cook Islands | 13 229 | 1 | Island mapping | Government population statistics | Direct contact and personal knowledge of members of collaborative working group | Data obtained from theatre logbook and/or anaesthetic registry of single bellwether hospital (no pre-existing centralised database available) | Single hospital data obtained from theatre registry and hospital mortality registry |
| Solomon Islands | 602 000 | 2 | Island mapping | Government population statistics | Direct contact and personal knowledge of members of collaborative working group | Data obtained from theatre logbook and/or anaesthetic registry from each bellwether hospital (no pre-existing centralised database available) | Mortality data obtained from both bellwether hospitals by direct contact No pre-existing reporting system for POMR |
| Nauru | 10 084 | 1 | Island mapping Single island with entire population within 30 min of hospital | Government population statistics | Direct contact and personal knowledge of members of collaborative working group | Data obtained from theatre logbook and/or anaesthetic registry of single bellwether hospital (no pre-existing centralised database available) | Single hospital data obtained from theatre registry and hospital death certificates |
| PNG | 8 000 000 | 30 | Non-mapping technique Local clinicians used locations of bellwether hospitals and adjacent regions within 2 hours combined with regional population data from Bureau of Statistics to calculate population within the 2-hour zone. | Government population statistics | Member of working group had direct contact with regional directors and annual provincial surgical reports submitted to Ministry of Health. No centralised database available | Data obtained from theatre logbook and/or anaesthetic registry from each provincial bellwether hospital The capital Port Moresby General Hospital data obtained directly from theatre registry and data from 29 referral hospitals obtained through recently presented PNG annual symposium Annual surgical reports, surgical audits, and direct email and telephone contact (no pre-existing centralised database available) | Never previously collected Mortality data from hospital death certificates at major referral hospital (Port Moresby) and audits from regional hospitals collected and presented for a recent surgical symposium |
| New Zealand | 4 452 300 | 20 | Manual contour line | Government population statistics | Medical Workforce Registration Database available from New Zealand Medical Council | Pre-existing centralised database maintained by government/Ministry of Health The calculation uses number of admissions where a General anaesthesia is administered as the measure of surgical volume; hence, two errors are acknowledged but thought to be small (<2% of patients have more than one operation in a single admission and regional anaesthesia without any GA administration is relatively uncommon). Endoscopy is not included in these data. | National Perioperative Review Committee annual reports |
| Micronesia | 102 109 | Island mapping | Government population statistics | Direct contact with Ministry of Health and clinical directors through working group | Not available No pre-existing centralised database available | Not available No pre-existing centralised database available | |
| Kiribati | 110 000 | 3 | Island mapping | Government population statistics | Direct contact with Ministry of Health and clinical directors through working group | Data obtained from theatre logbook and/or anaesthetic registry of single bellwether hospital (no pre-existing centralised database available) | Mortality data obtained from the three bellwether hospitals by direct contact using hospital death registry and personal logbooks of surgeons No pre-existing reporting system for POMR |
| Tuvalu | 10 800 | 1 | Island mapping | Government population statistics | Direct contact with Ministry of Health and clinical directors through working group | Data obtained from theatre logbook and/or anaesthetic registry of single bellwether hospital (no pre-existing centralised database available) | Single hospital data obtained from theatre registry and hospital mortality registry |
Definition of terms:
200 km radius zone: In Australia, where the road infrastructure in remote areas permits ambulance travel at 100 km/hour or more, a 200 km radius around each bellwether hospital was mapped, and cities or towns outside these areas were identified and populations estimated from population data available from the Australian government Bureau of Statistics.
Manual contour line: contour line determined by local clinicians with knowledge of local geography and travel conditions.
Island mapping: For archipelago nations comprising population distributed over multiple small islands, the 2-hour access zone included the island of the bellwether hospital (if access to the hospital within that island was less than 2 hours for all regions) and any island where the travel time was less than 60 min by locally available boat (or plane if readily available), which allowed for the complexity and time of transportation in these circumstances.
Population distribution data: government population statistics: government Bureau of Statistics population from latest census data using smallest available regions/towns/islands.
Where a region or island was not entirely within the contour line, a binary decision was applied to the population of the entire region based on a visual estimate of the percentage of land area contained within the 2-hour access contour line. If it appeared that 50% or greater of the land area of the region was within the contour line, then this region’s population was included within the <2 hour access group; otherwise the population was considered outside the 2-hour access zone.
SAO density: the number of SAOs working in a country was estimated using either a medical workforce database (necessary and available in Australia, New Zealand and Fiji) or direct contact with the Ministry of Health or clinical directors by members of the working group. In nations where the number of SAOs was less than 20, senior clinicians knew each SAO, and clinical directors in bellwether hospitals were able to confirm current levels of SAO workforce.
Surgical volume: either pre-existing centralised database maintained by government/Ministry of Health or data obtained from theatre and/or anaesthetic registry of each bellwether hospital where no pre-existing centralised database was available.
PNG, Papua New Guinea; POMR, perioperative mortality rate; SAO, Surgical, Anaesthesia and Obstetrics specialists.