Economic level is one of the most important factors for national health-care workforce and service distribution. China has been ranked as an upper middle income country by the World Bank since 2011. With the economy growth, China's health workforce (including anaesthesiologist) has been expanded (Figure 2). From 2012 to 2014, the GDP per capita in China increased about 18.0% and the density of physicians increased about 9.3% (3). The slower growth in density of physicians than in economy may explain why China's density of total physicians (14.9 per 10,000) is close to, but still lower than the average level (16.1 per 10,000) of upper middle income countries in the world. The global density of anaesthesiologists, obstetricians and surgeons was reported significantly correlated with the country's economic level (10). This survey found that the density of anaesthesiologists in China was 5.7 per 100,000 population in 2014, higher than those in lower middle income and low income countries, but much lower than those in high income countries, such as France (14.8), USA (12.7), UK (11.5), Israel (10.8) and Korea (9.2, Table 1). Economy influence on the distribution of health-care workforce and service was also found within China. GDP per capita in different provinces is positively correlated with the densities of physicians, anaesthesiologists and anaesthesia cases. The differences in the densities of physicians (highest in Beijing as 35.0/10,000 population vs. lowest in Tibet as 12.5/10,000 population) and anaesthesiologists (highest in Beijing as 12.6/100,000 population vs. lowest in Tibet as 2.9/100,000 population), as well as anaesthesia cases (highest in Beijing as 5788/100,000 population vs. lowest in Shanxi 1393/100,000 population) between different provinces varied by factors of 2.8 or higher, revealing severe unjust distribution of medical resources and anaesthesia service in China. Low density of health workers was associated with low medical coverage and high medical mortality in the world (18). Such huge un-equity distributions, consistent with previously reported for China (19), strongly suggested that China's central government should take some action to achieve fairness in health-care workforce and service distribution for the people living in different areas, but within the same country. Indeed, the health-care reform pushed by the National Health and Family Planning Commission will help China to improve such within-country equity (18).
Workload disparity of anaesthesiologists and surgeons among different sized hospitals in China was found in this survey. Anaesthesia workload among different provinces varied by a factor of 2.2 (Figure 3). In higher-level and larger-size hospitals, both inside and outside OR anaesthesia cases increased faster from 2012 to 2014, one attending anaesthesiologist did more anaesthesia cases and one attending surgeon operated on more patients a year (Table 2-4). These findings indicated that patients in China were more intended to visit doctors in larger hospitals, which led to the doctors' higher workload. Taken anaesthesiologists as an example, one attending in largest 10 hospitals in ARDC did 1433 anaesthesia cases in 2014 (1060 inside OR plus 373 outside OR), which was 67% more inside OR anaesthesia cases and 42% more outside OR anaesthesia cases than the national average level (Table 2). Chinese patients and the public trust in physicians in larger hospitals, rather than smaller hospitals, even for common and frequently-occurring diseases. This may be mainly due to the absent of nationwide standardised resident training (SRT) and hierarchical medical system in China's history. Training and quality of doctors in China is highly variable. As we found in this survey, 11549 of 34273 anaesthesia residents (only 33.7%, Table 2) were taking clinical training in the largest 1000 hospitals, the other 2/3 anaesthesia residents were doing their residency in smaller-size and lower-level hospitals. Actually, same situation exists for all other medical specialties in China. Right after medical school, most medical graduates were permanently employed by lower-level and smaller-size hospitals, such as township, county or small prefectural hospitals, and worked in the same hospital for entire career life in a way similar to civil servants. Even though called as residents in the first 5 years, they were not well trained during their residency in those hospitals to have necessary medical knowledge, clinical skills, ethical and humanistic professional competencies. Therefore, they are not able to provide same quality medical care for common and frequently-occurring diseases and to have patients' or public trust as who worked in higher-level and larger-size hospitals. Besides, health referral system was also weak in China, patients seeking high-quality primary health-care flooded directly into large hospitals, resulting much heavier workload for the physicians there.
Medical reform and education transformation in China have been aimed to meet the needs from Chinese people (20, 21). One of priorities in China's medical reform is to establish a nationwide SRT system. At the end of 2013, seven ministries of the central government jointly launched the SRT for all medical specialties including family medicine as a national strategy (22, 23). This SRT started in 2015 and will become compulsory by 2020. To produce competent doctors, all medical graduates are required to be trained in one of medical specialties including family medicine for three years in qualified training bases, most of which are large prefectural, provincial, and national hospitals. Nowadays, 559 training bases have been accredited across China by Chinese Medical Doctor Association and accredited bases will be expanded to about 1000, mainly match with the largest 1000 hospitals in ARDC in this survey. Since almost all residents will be trained in these largest 1000 hospitals, faculty size should have to keep the pace for resident growth in future. Considering more teaching, more complex and difficult cases, and more academic research conducted by attending faculties in training bases, the largest 1000 hospitals in ARDC should strengthen their faculties (number and quality) rapidly during the establishment phase of SRT from 2015 to 2020. The attending to resident ratio in the largest 1000 hospitals should reach 2-3 to 1, rather than 1.35 to 1 (15639 to 11549) for anaesthesiologoy in the largest 1000 hospitals found in ARDC of this survey. Standardised training during residency is fundamental for standardised practice for a physician at any level and sized hospital, and standardise practice is essential prerequisite and foundation for hierarchical medical system. Only when health-care provided for common and frequently-occurring diseases in primary hospitals were similar as larger hospitals, hierarchical medical system could be promoted and accepted well by patients and public. Another challenge that China has to overcome is to find a sustainable way in the near future to motivate these better-trained doctors to work at the grassroots level hospitals in rural and remote areas, and to reduce the brain drain from undeveloped regions.
To assess a nation's physician supply should not only assess the overall density of physicians, but also the infrastructure and distribution among different specialties. Disparity between surgeons and anaesthesiologists in China was found from this survey. Anaesthesiologists' shortage and surgeons' surplus were severer in higher-level and larger-size hospitals. According to a recent survey from 167 countries (including China) representing 92% of the global population, there were 1,112,727 surgeons, 483,357 obstetricians, and 550,134 anaesthesiologists (24). The ratio of surgeons plus obstetricians to anaesthesiologists was 2.9 to 1. Compared with this global ratio, China had a much higher ratio as 7.5 to 1 in 2014 (Table 4). In China, the larger the hospital was, the larger the ratio was. Attending surgeons to attending anaesthesiologists ratio in the largest 10 hospitals in SRDC was almost 10 to 1. Such relative surplus of surgeons was mirrored by that one attending surgeon in China had less than 1 operation day per week and performed only 86 operations in 2014. Again, the larger the hospital was, the less the operation days per week for one attending surgeon. In the largest 1000, 100 or 10 hospitals in SRDC, one attending surgeon had only 0.69, 0.64 or 0.58 operation days per week, respectively. The disparity of surgeons to anaesthesiologists ratio restricts multidisciplinary teamwork for surgical operations.
Actually, severe misdistribution does not exist only between surgeons and anaesthesiologists, but also exists in other specialties in China, such as family doctors, paediatricians and psychiatrists (25-28). Historically, China did not have a national needs-based resident recruitment system to keep a reasonable balance between different specialties. Except for personal willingness, due to unequal working condition, unbalanced payment and uneven social recognition, the number of medical graduates to apply residency in different specialties varies considerably. Adding to the lack of needs-based resident recruitment system, the current phenomena for application and recruitment in different specialties in China could be described as "more applied, more recruited, less applied, less recruited". In some places, even "no applied, no recruited" for paediatricians occurred (29). Since SRT will become compulsory, the preexisting imbalance between different specialties is possible to be exacerbated without a paralleling establishment of needs-based resident recruitment system. China's government is very strong and will be the leverage point for solving these problems. Set-up phase of compulsory SRT from 2015 to 2020 will be a good opportunity for government to set up a national needs-based resident recruitment system for optimizing residents allocation today and health-care workforce allocation in the future. The infrastructure of this system should include continued data collection and comprehensive assessment about the current shortage and surplus for all medical specialties, dynamic healthcare needs and demand-based analyses and projection of the future medical service needs and supply (30), allowing medical graduates to do multiple applications for different specialties, an effective recruitment match system for different economical regions, sized hospitals and medical specialties. China, the most populous country with about 20% of doctors in the world (27), recently completed a rapid transition from a low income country to an upper middle income country. China's experiences in medical reform, lessons on maldistribution and imbalance of healthcare workforce and workload should be, to some degree, valuable references for other developing countries those are making the transition from low or lower middle income countries to lower middle or upper middle income countries.
This was the first time to collect national data in mainland China, so we cannot compare the current situation to the past situation. We therefore cannot accurately describe the development condition of anesthesia workforce in mainland China in recent years.
In summary, mainland China had at least 77926 anaesthesiologists who covered 39.13 million anaesthesia cases in 2014. China has the biggest number of anaesthesiologists in the world, but is still in shortage of anaesthesia workforce. Significant maldistribution and imbalance in different provinces, sized hospitals and medical specialties were found in mainland China from this survey. We strongly suggest China's government to establish a nationwide needs-based resident recruitment system along with the set-up of compulsory SRT system.