Ambulance and Patient Transport Services

Ambulance and Patient Transport Services include Emergency Medical Services (EMS) and private ambulance services, which supply emergency prehospital care, including basic medical support and roadside transport to hospitals for patients experiencing medical emergencies. In recent years, a number of economists have written thoughtful and careful papers on EMS; this article will summarize their work and the work of others who write on EMS topics of interest to economists. Sections Taxonomy of Ambulance and Patient Transport Services and US Emergency Medical Services contain an introduction to EMS. Section Private Provision of Emergency Medical Services, describes the work analyzing the decision to outsource EMS. Section Factors Affecting Quality of Care, summarizes the existing evidence on supply side factors affecting the quality of care. Section Quality of Care and Health Outcomes describes research on the relationship between quality of care and health outcomes. Section Demand for Emergency Medical Services explores factors predicting demand. Section Cost-Effectiveness/Cost–Benefit Analyses includes a description of cost-effectiveness and cost–benefit analyses. Section Conclusion concludes.

Taxonomy Of Ambulance And Patient Transport Services

EMS are rival and excludable – only one patient can use an ambulance at a time and patients can be barred from service. In practice, however, access to EMS is frequently available to all, not only to those with the ability to pay. This makes EMS an impure local public good. EMS could also be considered an option good; patients frequently pay through taxes for the option of having EMS available when needed. As EMS systems are built to address urgent, unpredictable needs, there may be excess capacity most of the time.

EMS systems vary tremendously throughout the world. In Japan, EMS are provided by Emergency Life Support Technicians who have limited roles – they can provide cardiopulmonary resuscitation (CPR), defibrillate patients, and insert an airway, but are prohibited from distributing drugs. In Germany, EMS is regulated and organized at the ‘Lander’ or state level; the German population has the right and guarantee of prehospital emergency medical care either through a physician available through 24 h house call or via EMS. A person calling for EMS in Germany would reach a central dispatcher and then, most likely, would be served by a two-tiered system including a physician-staffed ALS. In 1998, EMS in Russia was two-tiered and staffed by physicians, with nurses dispatching ambulances from a central location and treatment initiated in the field; many patients are treated by physicians and then not transported to the hospital, unlike in the US system, for example, where transport is required for compensation. In general, European prehospital care is more likely to include care from a physician or a nurse in addition to a paramedic compared with American ambulances that do not have personnel with more than paramedic training; for a description, by country, of prehospital care arrangements (organization of EMS system, ambulance staffing, and helicopter availability), see Lethbridge (2009).

In low and middle-income countries, prehospital care is frequently unavailable; if it exists, it is concentrated in urban areas, likely to be privately provided (and only available to those with the ability to pay), of uneven quality and largely unregulated, even though trauma, particularly as a result of car accidents, represents an increasingly significant and growing source of disability and mortality in developing countries. In Islamabad, Pakistan, police officers as well as physicians staff ambulances provided through a public–private partnership; members of the community, including Non-Government Organizations subsidize physician salaries, equipment, and ongoing operational costs other than police salaries. In Turkey in the late 1990s, no personnel or equipment standards existed for prehospital care; in a typical city, Izmir, ambulances were staffed with a medical doctor with limited training and a driver without medical expertise, and it was unusual when the ambulance arrived before the patient had been transported by other means to the hospital. In 1997, Vietnam had no organized prehospital system; ambulances may be used for transport, but most often prehospital care relied on bystanders’ transporting patients. In 1998, consistent with many other developing countries, there was no centralized prehospital care system in Thailand; approximately 30 pick-up trucks staffed by volunteers picked up residents around Bangkok. Drivers have limited first-aid training. A water rescue boat must travel first from the hospital to the river, decreasing its usefulness significantly. Despite a large and increasing number of traffic accidents, prehospital care in India is largely nonexistent; with no centralized regulating body and the ambulance services only provided in only a few large cities where they are largely privately funded, most Indians lack access to trauma care of any kind. What is provided is of uneven quality; few programs exist to train paramedics and Emergency Medical Technicians (EMTs), and no certification or accreditation exists for professionals or programs. These characteristics define prehospital care throughout Southeast Asia (Bangladesh, India, Nepal, Pakistan, Bhutan, Maldives, and Sri Lanka); disproportionately concentrated in urban areas, serving those of higher socioeconomic status, frequently privately provided, without regulation or certification requirements, and limited in capabilities.

US Emergency Medical Services

In a typical EMS call in the US, a patient calls 911. A dispatcher at a local call center asks the patient a series of questions, evaluating the situation and eliminating false calls. The dispatcher may also give the patient medical instructions over the phone while simultaneously activating the local EMS response. In urban areas, first responders typically arrive first at the scene. A first responder captures vital signs, determines the patient’s medical history, and provides CPR. Meanwhile, the EMS response team composed of basic or intermediate EMTs or paramedics (advanced EMTs) travels to the scene by helicopter or by ground ambulance. Although the particular responsibilities of each type of personnel differ by state, EMTs supply more advanced care to patients than first-aid trained first responders. After arriving at the scene, assessing the situation, and providing initial care, the EMT or paramedic loads the patient into an ambulance or a helicopter and takes the patient to a hospital. In some cases, a medical director instructs and authorizes treatments en route. After transferring the patient to the care of physicians within the hospital, the EMS personnel collect billing information and fill out a call log with demographic and incident characteristics. In rural areas, the ambulance would likely be staffed by volunteers capable of delivering Basic Life Support services.

Most large cities in the US publicly provide EMS; in nearly half of all communities, EMS are organized and delivered through the fire department. Although first responders are almost always employed by a local government, either public or private ambulance or helicopter services may transfer patients. Many communities outsource emergency transport to for-profit ambulance agencies (more than 3000 in the US) or to hospital-based companies (approximately 7% of systems). In a hospital-based EMS system, the ambulance might park at the hospital in between calls and might be encouraged to bring patients to the affiliated hospital. With a private agency (hospital-based or other), the provider would likely own the infrastructure including the ambulance.

Revenues collected from private and public insurance for patient transports provide the majority of funding for EMS, potentially encouraging agencies to transport patients for whom the trip to the hospital is unnecessary. State and local taxes frequently supplement fees collected through insurance, along with grants from the state and the federal government. A variety of mechanisms, including government grants, fundraising, and donations, fund volunteer ambulance services.

Rather than being transported by ground ambulance, some patients may travel by medical helicopters. As of 2006, more than 650 medical helicopters operated within the US, run by private for-profit providers, hospitals, government agencies, or the military. More expensive to operate than traditional ambulances, helicopters may be no faster than ground ambulances, except in rural areas far from hospitals or in places where a ground ambulance cannot travel. Many patients transported by helicopter could have safely been transported by ground ambulance at considerably less expense without any survival loss. Using a helicopter may also limit the set of hospitals that a patient can be transported to.

Private Provision Of Emergency Medical Services

When do some communities choose to outsource patient transport? In a 2009 paper, Holian hypothesizes that a vote maximizing politician will outsource patient transport when it will increase her votes. In his model of private provision, as the proportion of the elderly, who consume a disproportionate amount of EMS rises, service levels change. Empirical work suggests an inverted U-shaped relationship between the proportion of the voting population which is elderly and the proportion of privately provided ambulance services.

Communities might outsource their EMS for many reasons. In 2009, David and Chiang found that although fire departments may have lower EMS transportation costs because they can take advantage of the existing firehouse infrastructure to get closer to patients, it may be cheaper for private agencies, which can spread costs across multiple communities – to introduce technology which improves the quality of care (such as Geographic Information System). Arguably, then, the decision to privatize depends on several factors including the distance to other cities, the population, and the number of hospitals in the city (all but the former negatively associated with private provision). Among the ten largest and ten smallest cities in the US, larger cities, with older, less healthy populations, a higher chance of disasters, more crime, less geographically dispersed fire stations and trauma centers, and strong unions, tend to be less likely to contract with private providers.

A related question not yet evaluated empirically is whether public or private agencies are better providers. Some hypothesize that private ambulances may provide EMS care more efficiently than public ambulances, because private paramedics frequently earn lower salaries than paramedics employed directly by state and local governments, even as they appear to have more sophisticated equipment and greater flexibility.

Factors Affecting Quality Of Care

Unfortunately, there are no nationally or internationally agreed-upon measures of EMS quality. However, response time, defined as the difference between the time of the initial call and the time of arrival at the scene, is one commonly used metric. Other metrics commonly used include total call time. Such metrics have not been systematically used by communities or states in the US to assess the quality of their EMS because a large proportion of states do not systematically collect response time data.

Many factors appear to be correlated with response times. In one southern state, Mississippi, whites appear to have higher response times than blacks, but these differences are eliminated after controlling for a county-level measure of population density. Others have found that distance, evening rush hour, patient being of Native American or Pacific Islander race, and gender predict longer total response times and that these factors plus bypass, neighborhood population density and percentage of white population are associated with delays of more than 15 min. Other factors including population density, the age of the housing stock, per-capita income, and first responders per square mile seem to be negatively correlated with mean response times, with area being positively correlated with mean response times.

It appears that incentives also affect response times – or at least the reported response times. One program in England profiled by Bevan and Hamblin (2009) publicly rewarded agencies meeting response time targets with gold stars. After the program was introduced, the proportion of agencies meeting performance targets increased, but the gains were illusory – response times were systematically shaved and calls recategorized as less severe to satisfy requirements.

Worker fatigue, experience, human capital depreciation, and turnover also affect response times. In a 2009 article, David and Brachet used the detailed call level data from Mississippi to measure the relationship between experience and time out of hospital or at the scene. They construct person-specific and firm-specific measures of experience, and control for individual fixed effects and a lengthy set of covariates. A one standard deviation increase in the number of trauma runs conducted by an individual in a given quarter is associated with a reduction of 35 s in out-of-hospital time and 10 s on scene. Brachet et al. (2010) compare the performance of paramedics working late at night in 24 h shifts with those same paramedics working late at night on 12 h shifts. They observed that paramedics on 24 h shifts have significantly longer response times and take longer to transport patients to the hospital and perform fewer procedures. David and Brachet’s (2011) article uses incident level data to measure the impact of human capital depreciation and turnover on time out of hospital. Turnover among EMS personnel is a significant problem for all EMS agencies, both paid staff and volunteers; one estimate puts the annual turnover among EMS personnel as high as 10%, with a median cost to agencies of over US$70 000. Partitioning experience into the human capital of those who work at the firm, those who have left the firm, and those who are joining the firm; David and Brachet derived an expression for firm-level experience and construct a measure of the relative contribution of turnover and human capital depreciation to organizational forgetting. Their reduced form estimates of organizational forgetting suggest that a quarter of the stock of experience existing at the beginning of the year survives to the end. When experience is separated into human capital accrued by individuals in the firm and those who have left the firm, they find the turnover to be a larger source of organizational forgetting (twice as large) than human capital depreciation.

Quality Of Care And Health Outcomes

How do factors which affect response time affect health? Although there are many studies that look at factors that affect the quality of EMS care, few evaluate the relationship between quality of care and health outcomes largely because of the challenges in linking prehospital records to mortality and hospital records and in finding a credible nonexperimental identification strategies in a context where experiments may not be feasible.

Athey and Stern’s (2002) work uses a differencesin-differences approach to determine the impact of the introduction of the new 911 technology on health outcomes. They model health as a function of response time and initial incident severity; they find that the introduction of Enhanced 911 in Pennsylvania improves the intermediate health measures for patients suffering from cardiac emergencies, as well as improving mortality measured 6 and 48 h after the initial incident. Enhanced 911 also reduces hospital costs for cardiac emergency patients. Wilde takes a different approach in her 2008 paper; she uses distance to the closest EMS agency as an instrument for EMS response time to account for the potential endogeneity of response time to patient severity. She finds that response time matters for mortality, but not health care utilization.

Shen and Hsia investigated the impact of bypass or diversion by EMS providers on mortality after acute myocardial infarction in a 2011 JAMA paper – an event which is arguably unrelated to the characteristics of the patient. Diversion may affect outcomes by affecting EMS response times (when the nearest hospital is on diversion, patients must be transported to hospitals that are farther away); it may mean that patients are transported to poorer quality hospitals or hospitals less capable of providing adequate care; it may also be an indicator for the quality of care for patients within the hospital experiencing the diversion (more crowded hospitals may provide worse care). Patients whose closest emergency department is on diversion for more than 12 h on the day of the incident experience higher mortality 30 days, 90 days, 9 months, and 1 year after the initial incident.

An example of work that explores a key policy question in EMS without a natural experiment or randomized controlled trial is that of a 2008 work by Concannon et al. who conducted a simulation of different EMS treatment choices for patients with acute ST-segment elevation myocardial infarctions. Patients can either be transported to the closest available hospital, transported only to hospitals with the capability of providing primary percutaneous coronary intervention (PCI) and treated with PCI or thrombolytic therapy, or be evaluated by EMS or by personnel at the local thrombolytic therapy-only hospital and then transported for PCI. Concannon et al. observed that selecting high-benefit patients for transport to PCI-capable hospitals reduces mortality without major shifts in hospital volumes.

Demand For Emergency Medical Services

What affects the use of EMS? There appears to be distinct EMS usage patterns by day (more calls between 10.00 a.m. and 8.00 p.m.) and the day of week (more calls on Friday and Saturday). Age and race/ethnicity also predict usage: people over the age of 85 years call more than 3 times the rate of those between 45 and 64 years of age and are transported at more than 4 times the rate of patients between 45 and 64 years of age. African Americans also call at a much higher rate than non-Hispanic whites.

In an intriguing analysis, Ringburg et al. conducted a discrete choice experiment in the Netherlands and found that households were willing to pay much higher amounts than would actually be necessary to provide 24 h helicopter emergency medical service as described in a 2009 paper. It appears that even if helicopter services are not cost-effective, households are willing to pay for them.

Many researchers in the field of operations research and applied mathematics have tackled questions regarding the optimal design of EMS systems, including identifying the optimal ambulance and helicopter station location and the optimal response time threshold for performance measurement, in addition to building models to forecast demand. That research is beyond the scope of this work.

Cost-Effectiveness/Cost–Benefit Analyses

Most existing cost analyses compare the costs and benefits of particular intervention or mode of care. For example, in their 2002 paper Athey and Stern calculate the costs and benefits from introducing Enhanced 911, a service that helps dispatchers to identify caller locations. They find that improvements in outcomes for cardiac issues cover 85% of the costs of implementing Enhanced 911, making the policy almost certain to be beneficial. Wilde conducts a cost–benefit analysis of a reduction in response times caused by eliminating mutual aid – a policy whereby communities share resources to cover excess demand – and finds that the per life year cost of a 9.5 s reduction in response times would be considerably less than US$50 000.

Evidence on the cost-effectiveness of air transport is mixed. One study that determined the costs of operating a local air ambulance service, supplemented with hospital costs for trauma survivors, estimated the cost of air transport per life year saved as US$2454. Another study collected microlevel costs, surveyed patients two years after their initial trauma incident, and estimated the incremental cost per quality-adjusted life-year (QALY) of helicopter use at more than 28 000 Euros. Several other studies looked retrospectively at patient records and concluded that there were few benefits for patients from air transport, and considerable costs to the health care system. Unfortunately, many of these studies fail to identify the perspective (societal or other), the year the costs were gathered in, fail to include comprehensive costs, and are inconsistent in their assessment of effectiveness making it difficult to draw concrete conclusions (QALY or mortality).

Conclusion

In recent years, there has been an increase in the literature written by or for economists on EMS. Nevertheless, many key clinical and policy questions remain unanswered, providing scope for further research. Economists have much to offer in the field of EMS: by asking different types of questions (i.e., on private vs. public provision, or cost-effectiveness) and using different techniques. Given the growing recognition of EMS as an essential part of emergency care, such research should only increase in the coming years.

References:

1. Athey, S. and Stern, S. (2002). The impact of information technology on emergency health care outcomes. RAND Journal of Economics 33(3), 399–432.
2. Bevan, G. and Hamblin, R. (2009). Hitting and missing targets by ambulance services for emergency calls: Effects of different systems of performance measurement within the UK. Journal of the Royal Statistical Society: Series A (Statistics in Society) 172(1), 161–190.
3. Brachet, T., David, G. and Duseja, R. (2010). The effect of shift structure on performance: The role of fatigue for paramedics. NBER Working Paper 16418. Available at: http://www.nber.org/papers/w16418 (accessed 11.06.13).
4. David, G. and Brachet, T. (2011). On the determinants of organizational forgetting. American Economic Journal: Microeconomics 3(3), 100–123.
5. Lethbridge, J. (2009). Privatisation of ambulance, emergency and firefighting services in Europe – A growing threat? pp. 1–21. Report Commissioned by European Federation of Public Service Unions. Available at: http:// www.psiru.org./ (accessed 05.06.10).
6. Concannon, T. W., Griffith, J. L., Kent, D. M., et al. (2009). Elapsed time in emergency medical services for patients with cardiac complaints are some patients at greater risk for delay? Circulation: Cardiovascular Quality and Outcomes 2(1), 9–15.
7. Concannon, T. W., Kent, D. M., Normand, S. L., et al. (2008). A geospatial analysis of emergency transport and inter-hospital transfer in ST-segment elevation myocardial infarction. American Journal of Cardiology 101(1), 69–74.
8. David, G. and Brachet, T. (2009). Retention, learning by doing, and performance in emergency medical services. Health Services Research 44(3), 902–925.
9. David, G. and Chiang, A. J. (2009). The determinants of public versus private provision of emergency medical services. International Journal of Industrial Organization 27(2), 312–319.
10. David, G. and Harrington, S. (2010). Population density and racial differences in the performance of Emergency Medical Services. Journal of Health Economics 29(4), 603–615.
11. Holian, M. J. (2009). Outsourcing in US cities, ambulances and elderly voters. Public Choice 141(3–4), 421–445.
12. Institute of Medicine (US). Committee on the Future of Emergency Care in the United States Health System (2007). Emergency medical services at the crossroads. Washington, DC: National Academies Press.
13. McConnel, C. E. and Wilson, R. W. (1998). The demand for prehospital emergency services in an aging society. Social Science & Medicine 46(8), 1027–1031.
14. Ringburg, A. N., Buljac, M., Stolk, E. A., et al. (2009). Willingness to pay for lives saved by helicopter emergency medical services. Prehospital Emergency Care 13(1), 37–43.
15. Shen, Y.-C. and Hsia, R. Y. (2011). Association between ambulance diversion and survival among patients with acute myocardial infarction. Journal of the American Medical Association 305(23), 2440–2447.
16. Wilde, E. (2008). Do response times matter? The impact of EMS response times on health outcomes. Princeton University Industrial Relations Section Working Paper.