Article, Emergency Medicine

Sustainable emergency medical service systems: how much energy do we need?

a b s t r a c t

Objective: Modern emergency medical service (EMS) systems are vulnerable to both rising energy prices and po- tential energy shortages. Ensuring the sustainability of EMS systems requires an empirical understanding of the total energy requirements of EMS operations. This study was undertaken to determine the life cycle energy re- quirements of US EMS systems.

Methods: Input-output-based energy requirement multipliers for the US economy were applied to the annual budgets for a random sample of 19 metropolitan or county-wide EMS systems. Calculated per capita energy re- quirements of the EMS systems were used to estimate nationwide EMS energy requirements, and the leading en- ergy sinks of the EMS supply chain were determined.

Results: Total US EMS-related energy requirements are estimated at 30 to 60 petajoules (1015 J) annually. Direct (“scope 1”) energy consumption, primarily in the form of vehicle fuels but also in the form of natural gas and heating oil, accounts for 49% of all EMS-related energy requirements. The energy supply chain–including system electricity consumption (“scope 2”) as well as the upstream (“scope 3”) energy required to generate and distrib- ute liquid fuels and natural gas–accounts for 18% of EMS energy requirements. Scope 3 energy consumption in the materials supply chain accounts for 33% of EMS energy requirements. Vehicle purchases, leases, maintenance, and repair are the most energy-intense components of the non-energy EMS supply chain (23%), followed by medical supplies and equipment (21%).

Conclusion: Although less energy intense than other aspects of the US healthcare system, ground EMS systems require substantial amounts of energy each year.

(C) 2014

Introduction

Concerns about the energy consumption necessary to sustain health services [1-4], and the environmental impact of that energy consump- tion [5-9] continue to increase. Energy is a critical input for all health services [1-4,10-12], necessary for operating health facilities, producing and distributing medical supplies and equipment, and providing health- related transport including emergency medical services (EMS) [10,12]. Emergency medical service systems are a critical component of both the emergency medicine and public health infrastructure, but they are increasingly vulnerable to rising energy prices and potential energy shortages. Several media reports have described the financial challenges rising fuel prices present for the operating budgets of EMS systems in the United States, the United Kingdom, and Canada [13-17]. In Australia, increases in energy costs have been demonstrated to adverse- ly affect not only operational budgets but also EMS system performance including Response times and work-related injury rates [18].

? Presented in part at the National Association of EMS Physicians Annual Meeting, Tucson, Arizona, January 2014.

* Corresponding author at: Mt. Isa Centre for Rural and Remote Health, James Cook University, Townsville, QLD 4810 Australia.

E-mail address: [email protected] (L.H. Brown).

Currently, energy-related sustainability is not the most pressing issue facing most US EMS systems. Funding, staffing, training, quality improvement, and risk management are ever-pressing operational con- cerns for EMS systems [19], few of which are fully resourced to address those issues. At a policy level, however, ensuring the sustainability of EMS systems in the face of mounting energy insecurity and rising ener- gy prices is important for maintaining the integrity of the health care system. At a minimum, this requires an empirical understanding of the total energy requirements of EMS operations. This study was under- taken as an initial step in developing that understanding, with the spe- cific aims of (1) determining the life cycle energy requirements for a randomly selected sample of US EMS systems, (2) estimating the total annual life cycle energy requirements for EMS in the United States, and (3) identifying those items that contribute most to the energy de- mands of the EMS supply chain.

Methods

Characterizing and measuring energy consumption

Energy consumption can be subdivided into 3 “scopes” [20,21]. Scope 1 energy consumption is the direct consumption of fuels, such

http://dx.doi.org/10.1016/j.ajem.2014.11.011

0735-6757/(C) 2014

as diesel or gasoline in a motor vehicle. Scope 2 energy consumption re- lates to purchased energy, such as electricity, where the energy is gen- erated remotely from the location of use. Scope 3 energy consumption is a prorated portion of the energy consumed in a product’s or service’s supply and waste disposal chain, such as a share of the energy required to operate a factory that builds the truck chassis for the ambulances pur- chased by an EMS agency. Life cycle energy consumption is the sum of scope 1, scope 2, and scope 3 energy consumption [20,21]. Large-scale energy consumption is measured in multiples of joules, including megajoules (MJ, 106 J), gigajoules (GJ, 109 J), terajoules (TJ, 1012 J), petajoules (PJ, 1015 J), and exajoules (EJ, 1018 J).

Design

This study used a hybrid input-output-based approach to estimating the life cycle energy consumption of a convenience sample of randomly selected US EMS systems. Input-output analysis models transactions be- tween all of the sectors of an economy, including the inputs and by- products of those economic transactions [22]. Although input-output anal- ysis provides an estimation, not an accounting, of an entity’s energy and environmental burden, it is widely incorporated into studies of the energy consumption and environmental impacts of national economies [23,24], international trade [25,26], waste management strategies [27,28], global cities [29], individual firms [30,31], and health systems [7,8].

Input-output multipliers

The Green Design Institute at Carnegie-Mellon University has devel- oped energy requirement multipliers based on the input-output tables for the US economy [32]. The multipliers in their “purchaser price model” allow the estimation of the life cycle energy consumption asso- ciated with each dollar (in 2002 dollars) of goods or services purchased from any economic sector. For example, every $1 million of goods and services purchased from the “health care and social assistance” sector is associated with a total of 3.56 TJ of energy consumption in all of the upstream and downstream economic processes involved in delivering those goods and services [32]. The model includes multipliers for 428 different economic sectors, with descriptions of all of the various eco- nomic activities that contribute to each sector. Emergency ambulance operations are included in the health care and social assistance sector of the US economy; thus, the energy consumption of EMS systems is in- cluded in the multiplier for that sector [32].

Identifying the convenience sample

The US Census Bureau lists 289 cities with a population of 100 000 or more [33] and also lists all 3143 counties or county equivalents in the country [34]. Both of these lists were obtained, and a computerized ran- dom number generator (Excel “RAND,” Microsoft Corporation, Redmond, Washington) was then used to select 200 localities with roughly equal numbers coming from the city (n = 96) and county (n = 104) lists. We anticipated EMS budget information would be publicly available for ap- proximately 10% of the selected cities and counties, yielding an estimated convenience sample of 20 EMS systems.

The publicly available Web site of each selected city or county, where available, was accessed and searched to locate any publicly avail- able budget information for the local EMS system that could be differen- tiated from the budget for other activities (eg, Fire Department, Public Health Department). The most recent publicly available budget was saved for use in the primary analysis. The EMS budgets were then fur- ther evaluated to determine whether they contained detailed line item budget allocations. For those systems for which comprehensive itemized budgets were available, the 2 most recent budgets were saved for use in the secondary analyses. Preference was given to audited budgets detailing actual expenditures over proposed budgets with projected expenditures.

In addition to the budget data, each city or county’s estimated popu- lation data for the respective budget years were obtained from the US Census Bureau [33,34], and annual EMS system response volume data were obtained from publicly available annual reports.

Primary analysiscursory calculation of EMS life cycle energy requirements

Our primary analysis used a methodology similar to one used by Chung and Meltzer [8] to determine the greenhouse gas emissions of the US health care sector. All expenditures for the most recent budget year for each EMS system were transformed into year 2002 dollar amounts using consumer price index data [35]. These adjusted expendi- ture amounts were then multiplied by the health care and social assis- tance sector multiplier (3.56 TJ/$1 million of expenditure) [32] to calculate the annual life cycle energy requirements for each system. We then used the respective energy, population, and response volume data to determine the energy intensity of the EMS systems as GJ per 1000 population and GJ per 1000 ambulance responses. We report the range, median, interquartile range (IQR) for the calculated energy inten- sities. Each included EMS system’s per capita expenditures, the health care and social assistance sector energy multiplier, and the Census Bu- reau estimate of the 2012 total population (314 million) [36] were also used to estimate the annual EMS-related energy requirements for the entire Unite States (see Eq. (1)). We report the mean and 95% con- fidence interval (CI) for these 19 estimates.

[‘D’/’local pop’] x [‘energy/D1 million’] x [‘US pop’] (1)

where ‘$’ is the dollars expended; ‘local pop’ is the local population; ‘en- ergy/$1 million’ is the reported health care and social assistance sector energy requirements per $1 million of expenditure; and ‘US pop’ is the US population in millions.

Secondary analysisdetailed calculation of EMS life cycle energy requirements

Although the health care and social assistance multiplier is the techni- cally appropriate multiplier for estimating EMS energy consumption [32], modeling EMS energy requirements using that multiplier for all EMS ex- penditures assumes that the same amount of energy is required to pro- duce $1 worth of EMS service or $1 worth of inpatient care; it also assumes the same amount of energy is required to produce $1 worth of diesel fuel or $1 worth of laryngoscopes. To more accurately represent the specific energy requirements of EMS systems and the EMS supply chain, we also took a more detailed approach to the input-output analysis. For those EMS systems with detailed budgets, every budget line item was mapped by 1 investigator (LHB) to 1 or more of 46 relevant economic sectors incorporated into the Green Design Institute’s purchaser price model (see Appendix A), each with its unique energy requirements mul- tiplier [32]. Where a budget line item clearly combined expenditures in more than 1 economic sector, the expenditure was divided equally among the relevant sectors. For example, expenditures in a “medical sup- plies and pharmaceuticals” budget line item were divided equally be- tween the “surgical appliance and supplies manufacturing” and the “pharmaceutical preparation manufacturing” sectors. Employer-paid health, dental, and vision insurance premiums were allocated to the “insurance agencies, brokerages and related services” sector. Other tangi- ble employee benefits, such as health club memberships and clothing al- lowances, were allocated to the generic “retail trade” economic sector. No energy requirements were attributed to employee cash or cash-like com- pensation, including salaries, overtime, retirement contributions, unem-

ployment, and worker’s compensation payments.

Life cycle analysis provides an approximation, not an accounting, of energy demands. Thus, we made 3 different secondary calculations. First, purchases of liquid fuels, including gasoline, diesel, and fuel oil,

were allocated to the “petroleum refineries” economic sector. Next, be- cause EMS systems purchase fuels from wholesale or retail outlets, not directly from refineries, we repeated the secondary analysis using the less energy-intense retail trade multiplier for vehicle fuel and heating oil expenditures. Finally, because most EMS supply chain purchases in- volve some intermediary wholesaler and/or retailer, we used the retail trade multiplier for all purchases in all economic sectors.

As in the primary analysis, all expenditures were transformed into year 2002 dollar amounts, and the adjusted expenditure amounts with- in each economic sector were multiplied by the energy requirement multiplier for that sector. The resultant line item calculations of energy requirements were then summed to determine total indirect (scopes 2 and 3) energy requirements for each budget year for each EMS system. To determine scope 1 energy requirements, the volume of diesel, gasoline, heating oil, and natural gas purchased by each EMS agency was estimated by dividing the amount expended on each fuel type in a given budget year by the annual average price of that fuel for the same budget year as reported by the US Energy Information Administra- tion [37,38]. We then used volume-based energy multipliers from the Energy Information Administration [39] to determine the energy pro-

duction associated with the consumption of that volume of fuel.

We calculated life cycle energy requirements as the sum of the scope 1, scope 2, and scope 3 energy consumption for each budget year for each EMS system. We again used the respective energy, expenditure, population, and response volume data to determine the energy intensi-

Table 1

Characteristics of the included cities and counties

Location Population Annual EMS

budget

Systems with detailed budgets

Central

7500

$660,000

1200

Southeast

12 500

$1.2 million

3000

Northeast

17 500

$1.6 million

N/A

Southeast

27 000

$1.8 million

N/A

Southeast

41 000

$3.8 million

7000

Southeast

51 000

$2.8 million

N/A

Southeast

205 000

$15.0 million

51 000

Southwest

385 000

$13.9 million

30 000

Northeast

400 000

$24.3 million

90 000

Systems without detailed budgets

Central

19 000

$1.1 million

N/A

Central

27 500

$466,000

N/A

Southwest

120 000

$5.2 million

16 000

Northeast

125 000

$1.2 million

N/A

Southeast

230 000

$18.7 million

41 000

Southeast

280 000

$13.5 million

30 000

Southeast

355 000

$11.4 million

36 000

Northeast

625 000

$50.0 million

135 000

Southeast

630 000

$34.1 million

76 000

Southeast

945 000

$13.1 million

111 000

Abbreviation: N/A, not available.

Annual response volume

ty of EMS systems as TJ per $1 million dollars of expenditure, GJ per 1000 population, and GJ per 1000 ambulance responses. We also used the per capita expenditures of the 19 included systems, these 3 addi- tional expenditure-based energy intensities, and the total US population to again estimate EMS-related energy requirements for the entire Unite States, reporting the mean and 95% CI for those estimates.

Secondary analysisleading contributors to supply chain energy requirements

We also totalled all energy requirements across all budget years and all EMS systems for each of the included economic sectors. We ranked the sectors, from highest energy requirements to lowest, to identify those components of the supply chain that contribute most to the life cycle energy requirements of EMS systems.

Ethics

No human or animal subjects were involved in this study; all of the included data are publicly available. However, because of the potentially

sensitive nature of the data, we present only aggregated data and con- ceal the identities of the included EMS systems.

Results

The EMS budgets distinguishable from other services were available for 19 (9.5%) of the randomly selected cities or counties. Detailed line item budgets were available for 9 of the cities or counties; budgets for the most recent 2 years were available for 7 of these systems, whereas 2 cities provided only the most recent year’s detailed budget.

The 19 EMS systems included in the primary analysis were located in 12 different states in the northeastern (n = 4), southeastern (n = 10), central (n = 3), and southwestern (n = 2) continental Unite States. Their populations ranged between 7500 and 945 000, with annual response volumes ranging between 1200 and 135 000. The EMS system budgets ranged from $466 000 to $24.3 million annually, aver- aging $58 +- 27 per capita and $385 +- 120 per ambulance response. The included EMS systems are described more fully in Table 1, and the various estimates of their life cycle energy requirements are shown in Table 2.

Table 2

Life cycle energy intensities of the included EMS systems

Point estimate Minimum 25th percentile Median 75th percentile Maximum

“Health care” multiplier for all expenditures

TJ per $1 million

3.56 [32]

N/A

N/A

N/A

N/A

N/A

GJ per 1000 population

132.8

27.3

118.4

156.9

225.3

268.5

GJ per 1000 responses

913.5

323.5

885.8

1065.9

1295.1

1652.9

Petroleum” multiplier for fuel purchases TJ per $1 million 3.25

1.81

2.96

3.49

4.68

6.84

GJ per 1000 population

121.2

24.9

108.1

143.2

205.7

245.2

GJ per 1000 responses

833.9

295.4

808.7

973.1

1182.3

1508.9

RetailMultiplier for Fuel Purchases

TJ per $1 million

2.88

1.62

2.63

3.07

4.44

6.51

GJ per 1000 population

107.4

22.1

95.8

126.9

182.3

217.2

GJ per 1000 responses

739.0

261.7

716.6

862.3

1047.7

1337.2

“Retail” multiplier for all purchases

TJ per $1 million

2.63

0.98

2.36

2.79

3.42

4.73

GJ per 1000 population

98.1

20.2

87.4

115.9

166.4

198.4

GJ per 1000 responses

674.8

239.0

654.4

787.5

956.8

1221.1

Abbreviations: GJ, gigajoules; “Health care,” health care and social assistance sector; N/A, not applicable; “Petroleum,” petroleum and refineries sector; “Retail,” retail trade sector; TJ, terajoules.

Primary resultscursory calculation of EMS life cycle energy requirements

For the budget years included in the primary analysis, the 19 EMS systems spent $168 million (in 2002 dollars) delivering services to 4.5 million people. At 3.56 TJ per $1 million of expenditure [32], the com- bined annual energy consumption of these systems is calculated at 598 TJ, approximately 133 GJ per 1000 population (median, 156 GJ) or 914 GJ per 1000 responses (median, 1066 GJ).

Secondary analysisdetailed calculation of EMS life cycle energy requirements

The 9 EMS systems with detailed line item budgets spent $94.7 mil- lion (in 2002 dollars) over a combined 16 budget years. When using the sector-specific energy multipliers and applying the petroleum refineries multiplier to fuel purchases, these systems consumed roughly 307.9 TJ of energy for an overall energy intensity of 3.25 TJ per $1 million of ex- penditure (median, 3.49 TJ), approximately 121 GJ per 1000 population (median, 143 GJ) and 834 GJ per 1000 responses (median, 973 GJ).

Using the retail trade energy multiplier for all liquid fuel expendi- tures reduced total calculated energy requirements, for an aggregate en- ergy intensity of 2.88 TJ per $1 million of expenditure (median, 3.07 TJ), 107 GJ per 1000 population (median, 127 GJ), and 739 GJ per 1000 re- sponses (median, 862 GJ). Using the retail trade multiplier for all trans- actions further reduced calculated energy requirements, with an aggregate energy intensity of 2.63 TJ per $1 million of expenditure (me- dian, 2.79 TJ), 98 GJ per 1000 population (median, 116 GJ), and 675 GJ per 1000 responses (median, 786 GJ).

Estimated EMS-related energy consumption for the entire Unite States

Using the health care and social assistance multiplier and extrapolat- ing the per capita energy requirements of the 19 included EMS systems to the entire US population suggests that the EMS-related energy con- sumption in the Unite States totals 50 133 TJ (or 50.1 PJ; 95% CI, 39.8-60.4 PJ) annually. Using the energy intensity calculated in our sec- ondary analysis incorporating the petroleum refineries multiplier sug- gests that the total EMS-related life cycle energy requirements of approximately 45.8 PJ (95% CI, 36.4-55.2 PJ) annually, whereas esti- mates using the retail trade multiplier for fuel purchases (40.6 PJ; 95% CI, 32.2-48.9 PJ) or all purchases (37.0 PJ; 95% CI, 29.4-44.6 PJ) are

lower still. Figure shows the mean and 95% CIs for the estimates based on each of the 19 city or county budgets and the various combinations of energy multipliers considered in this study.

Leading contributors to supply chain energy requirements

Scope 1 energy consumption, primarily in the form of vehicle fuels but also in the form of natural gas and heating oil, accounted for 49% of all EMS-related energy requirements. The energy supply chain–in- cluding scope 2 electricity consumption and the scope 3 energy

required in the generation and distribution of liquid fuels and natural gas–accounted for 18% of EMS energy requirements. Scope 3 energy consumption in the non-energy EMS supply chain accounted for 33% of EMS energy requirements.

When focusing solely on the non-energy EMS supply chain, 5 expen- diture categories accounted for 75% of scope 3 energy requirements: ve- hicle purchases, leases, maintenance, and repairs (23%); medical supplies and equipment (20%); employee benefits (16%); office supplies and equipment (10%); and contracted and professional services (6%). The top 20 EMS supply chain energy sinks, accounting for 99.5% of sup- ply chain energy requirements, are shown in Table 3.

Discussion

These results highlight the energy intensity of EMS systems, and their potential vulnerability to energy shortages and price increases. The low end of the 95% CI for EMS-related energy consumption using our most conservative estimate was 30 PJ annually; the high end of the 95% CI for the least conservative estimate was 60 PJ annually. Al- though 30 PJ or even 60 PJ might seem a small proportion of the approx- imately 84 000 PJ (or 84 EJ) of energy produced in the Unite States each year [40], “emergency medical transportation services” is just 1 of more than 19 000 distinct economic activities that make up the US economy [41]. At that level of disaggregation, the energy requirements of any in- dividual activity appear small–on average, 4.4 PJ per economic activity (84 EJ/19 000). At 30 PJ of energy annually, EMS is 7 to 15 times more energy intense than average. For further context, 30 PJ of energy is ap- proximately the 2012 coal-fired energy production of New Hampshire, California, Delaware, Hawaii, Alaska, Connecticut, Idaho, and Maine combined [42]. Simply put, EMS requires substantial amounts of energy. These findings also reveal that vehicle fuel consumption accounts for only about half of the energy requirements of EMS systems, the other half manifesting in the energy and non-energy supply chains. Not only would persistent energy shortages or substantial price increases affect direct energy consumption and day-to-day EMS operations, but they would also have extensive indirect impacts on the availability and costs of vehicles, medical supplies, equipment, and other components of the EMS supply chain. These findings are consistent with previous EMS sustainability studies from Australia [18,43], and underscore the importance of proactively pursuing the long-term sustainability of EMS systems, finding efficiencies in all aspects of EMS operations, and creating contingency plans for mitigating fuel shortages and energy

price increases.

One promising finding of these analyses is that US EMS systems ap- pear to be slightly less energy intense, when measured as energy re- quirements per $1 million of expenditure, than some other aspects of the US health care system. The energy intensity calculated from our least conservative secondary analysis, 3.25 TJ per $1 million of expendi- ture (in 2002 dollars), is 9% less than the health care and social assis- tance sector multiplier of 3.56 TJ per $1 million of expenditure [32]. This might be partially explained by EMS systems not having the burden of supporting large, fixed health facilities such as hospitals, which are

Figure. Estimated annual life cycle energy requirements of US EMS systems (PJ) (mean, 95% CI). Abbreviations: PJ, petajoule; ‘Petroleum,’ petroleum and refineries sector; ‘Retail,’ retail trade sector.

Table 3

Leading contributors to non-energy EMS supply chain energy requirements

Rank

Item

% of supply chain energy

Cumulative % of supply chain energy

1

Vehicle lease/purchase, maintenance, and repair

22.8

22.8

2

Medical supplies and equipment

20.5

43.3

3

Employee benefits

15.8

59.1

4

Office supplies, stationery, equipment, and furniture

9.8

69.0

5

Contracted/professional/billing services

5.7

74.6

6

Nonmedical equipment/supplies and maintenance

4.7

79.3

7

Tenancy, leases, and building maintenance

3.9

83.2

8

Financial instruments, fees and interest, insurances

3.6

86.8

9

Telephone and Internet

2.2

89.0

10

Pharmaceutical supplies and medical gases

2.0

91.0

11

Uniforms and clothing

1.5

92.5

12

Shipping and postage

1.2

93.7

13

Advertising, marketing, printing, publishing

1.1

94.8

14

Radio equipment, supplies, and maintenance

1.1

95.9

15

Dues, registrations, training, and travel

1.0

96.9

16

Miscellaneous

0.7

97.6

17

Janitorial services and cleaning supplies

0.6

98.2

18

Health care and medical services

0.5

98.7

19

Laundry

0.5

99.2

20

Computers equipment, software, and support

0.3

99.5

known to be energy and environmentally intense [12]. This has potential implications for both energy and health policymakers. For example, net energy savings in the overall health sector might be an additional unantic- ipated benefit of innovative EMS initiatives that aim to reduce costs and improve patient outcomes by minimizing emergency department visits and hospital admissions. This is an important point about sustainability initiatives: they do not necessarily conflict with efforts to expand and im- prove service delivery, and can in fact successfully coexist with broader efforts aimed at financial and operational efficiency [44].

Clearly, improving the energy efficiency of the vehicle fleet and the building stock should be a priority for sustainability-minded EMS sys- tems, but EMS sustainability initiatives should not focus exclusively on ve- hicle fuels and electricity consumption. The non-energy supply chain is responsible for one-third of EMS life cycle energy requirements in the Unite States, with vehicles and medical supplies being the leading con- tributors to supply chain energy demands. As can be seen in Table 3, some of the leading contributors to EMS supply chain energy require- ments are intuitive; others are not. Employee benefits (16%), administra- tive and contracted activities such as billing services, financial services, and professional services (6%), and financial instruments, fees, and insur- ances (4%) are just some of the less obvious yet significant indirect energy sinks for US EMS systems. Optimizing the supply chain, minimizing waste, and pursuing efficiencies in administrative processes are all strate- gies for reducing the scope 3 energy requirements of EMS systems.

The contribution of the vehicle fleet to the energy requirements of EMS systems demonstrates the importance of taking a life cycle per- spective on energy consumption and sustainability efforts. Although an EMS system might reduce scope 1 energy consumption (and tailpipe greenhouse gas emissions) through an aggressive fleet modification ini- tiative, the scope 3 energy requirements associated with building and delivering those new vehicles must also be considered. Similar to a tra- ditional cost-benefit analysis, the reduction in life cycle energy con- sumption will depend largely on the vehicle lifespan and changes (if any) in maintenance requirements. Additional research will be neces- sary to determine the net energy benefits, if any, of fleet modification initiatives as well as other EMS sustainability efforts.

Limitations

This study is based on a small, randomly selected sample of ground EMS systems with publicly available budget information. Although our small sample size results in wide CIs for our estimates of energy con- sumption, there is no reason to think that more (or less) efficient EMS agencies would also be systematically more (or less) likely to post

budget information online, and that our sample is therefore not ade- quately representative of the target population. Indeed, EMS studies often include data for only large, urban, and suburban systems; we in- cluded data for small towns and counties to ensure a more complete representation of EMS in the Unite States. Although our study did not in- clude any of the 9 US cities with a population greater than 1 million, only about 7% of the US population live in such very large cities.

All of the localities in our study are served by municipal or public utility model EMS systems; we found no cities with publicly available EMS bud- gets from the western Unite States, where private proprietary EMS agen- cies are more common. It is conceivable that proprietary agencies have lower per capita or per response expenditures, and thus lower input- output-based energy intensities. It is unlikely that the distribution of ener- gy sinks for proprietary systems would differ substantively from those of the systems included in our analysis. Our analysis does not include fire- based EMS systems or systems where EMS budgets cannot be differentiat- ed from a parent department (eg, Public Safety) budget. Although there is no reason to think that the EMS component of these conglomerated sys- tems would be more or less efficient than the independent third-service EMS systems included in our analysis, there might be efficiencies in their administrative structures and purchasing processes that could lower their overall life cycle energy requirements. The effects of system structure on EMS system energy requirements and environmental burden is an area in need of further research. Finally, our analysis does not include data for air medical operations or independent, non-emergency transport agencies serving the same communities included in this analysis. The total energy burden of all EMS activities in the Unite States would be greater than the estimates for ground EMS calculated in this study.

We estimated, rather than measured, direct energy consumption by cal- culating the approximate volume of diesel, gasoline, heating oil, and natural gas consumed by the included EMS systems. A more exact approach would be to directly inventory EMS system fuel consumption. Interestingly, the scope 1 and scope 2 energy consumption calculated for the systems in this study (IQR, 0.35-0.69 TJ/1000 responses) was nearly identical to the scope 1 and scope 2 energy requirements found in a previous study, in which we did directly inventory the energy consumption of 14 completely different US EMS systems (IQR, 0.41-0.68 JT/1000 responses) [45].

We must also recognize the shortcomings of input-output analyses for estimating energy consumption. Evenly dividing the expenditures in some budget line items among the relevant economic sectors is inex- act, and might underestimate the energy requirements of some supply chain components and overestimate the requirements of others. In ad- dition, although we used sector-specific energy multipliers for each budget line item, these are still aggregate multipliers that assume all

expenditures within an economic sector result in the same amount of energy consumption. Importantly, these potential misclassifications and aggregations are mitigated in the analyses in which all expenditures were allocated to either the health care and social assistance or retail trade sectors. Finally, the input-output multipliers used in this analysis are based on the 2002 US economy. These are, however, the most recent publicly available input-output-based energy multipliers for the US economy, and the underlying structure of the economy has not changed significantly over the past decade. These are all common limitations of input-output-based life cycle analyses [30,46], including those in previ- ous evaluations of EMS and other health activities [7,8,43]. Despite these limitations, input-output analysis is a well-established and widely ac- cepted methodological approach to identifying and measuring the life cycle energy requirements of services and industries [23-31].

Conclusion

US ground EMS systems appear less energy intense than activities in the broader health care and social assistance sector, but still require sub- stantial amounts of energy to operate: at a minimum 30 PJ and as much as 60 PJ of energy annually. Improving the efficiency of the vehicle fleet and building stock is a clear priority for EMS systems, but the non- energy supply chain is responsible for one-third of EMS life cycle energy requirements and should not be forgotten in sustainability initiatives. Although these findings are specific to EMS systems, they can also in- form health and energy policymakers interested in the sustainability of health care systems more generally. We encourage further research to better quantify the energy requirements of EMS systems and other health care activities, to identify their environmental impacts, and to ex- plore and evaluate strategies for mitigating both.

Appendix A. Included economic sectors from the Green Design Institute’s purchaser price model [32]

Advertising & Related Services Other Commercial & Service Industry

Machinery Manufacturing

References

  1. Frumkin H, Hess J, Vindigni S. Peak petroleum and public health. JAMA 2007;298 (14):1688-90.
  2. Hess J, Bednarz D, Bae J, Pierce J. Petroleum and health care. Evaluating and manag- ing health care’s vulnerability to petroleum supply shifts. Am J Public Health 2011; 101:1568-79.
  3. Wilkinson P. Peak oil: threat, opportunity or phantom? Public Health 2008;122: 664-6.
  4. Hanlon P, McCartney G. Peak oil: will it be public health’s greatest challenge? Public

    Health 2008;122:647-52.

    Butler CD. Lightening our carbon footprint: economics, norms and doctors. Med J Aust 2010;192(9):485-6.

  5. Simpson M. Reducing NHS, carbon footprint: time for a culture change. BMJ 2008; 336:848.
  6. Sustainable Development Commission-Stockholm Environment Institute. NHS En- gland Carbon Emissions Carbon Footprinting Report. London, England: Sustainable Development Commission; 2008.
  7. Chung JW, Meltzer DO. Estimate of the carbon footprint of the US health care sector. JAMA 2009;302:1970-2.
  8. Nichols A, Richardson J. Climate change, health and sustainability: a brief survey of primary care trusts in the south west of England. Perspect Public Health 2011;131 (2):82-4.
  9. Frumkin H, Hess J, Vindigni S. Energy and public health: the challenge of peak petro- leum. Public Health Rep 2009;134:5-19.
  10. Schwartz BS, Parker CL, Hess J, Frumkin H. Public health and medicine in an age of energy scarcity: the case of petroleum. Am J Public Health 2011;101:1560-7.
  11. Brown LH, Buettner PG, Canyon DV. The energy burden and environmental impact of health services. Am J Public Health 2012;102:e76-82.
  12. Anonymous. Ambulance service’s budget pressure over diesel price. BBC News; 2011[Retrieved from www.bbc.co.uk on February 18, 2011].
  13. Fallon P, Va W. ambulance fuel costs over budget. Charleston [W.Va.] Daily Mail; 2011[Retrieved from www.ems1.com on July 25, 2011].
  14. Harlin K. Air ambulance company rebounds from tragedies, high fuel prices. Investor’s Business Daily; 2009[Retrieved from www.investors.com/ NewsAndAnalysis/Article.aspx?id = 506506 on December 7, 2010].
  15. Odell E. No more fuel for emergency vehicles. WAAY Television News; 2011 [Retrieved from www.waaytv.com/news on March 14, 2011].
  16. Penner DE. Rising fuel prices add millions to services costs. The Vancouver [BC] Sun; 2008[Retrieved from www.canada.com/vancouversun/news on September 23, 2010].
  17. Brown LH, Chaiechi T, Buettner PG, Canyon DV, Crawford JM, Judd J. Higher energy prices are associated with diminished resources, performance and safety in Australian ambulance systems. ANZ J Public Health 2013;37:83-9.
  18. Lerner EB, Nichol G, Spaite DW, Garrison HG, Maio RF. Comprehensive framework for determining the cost of an emergency medical services system. Ann Emerg Med 2007;49:304-13.
  19. Matthews HS, Hendrickson CT, Weber CL. The importance of carbon footprint esti-

    All Other Miscellaneous Professional & Technical Services

    Other Cut & Sew Apparel Manufacturing

    mation boundaries. Environ Sci Technol 2008;42(16):5839-42.

    Purman JR. Tracking your carbon footprint: a step-by-step guide to understanding

    Automotive Repair & Maintenance Other Educational Services Business Support Services Periodical Publishers

    and inventorying greenhouse gas emissions. New York, NY: iUniverse Inc.; 2008.

    Leontief W. Environmental repercussions and the economic structure: an input-out-

    Civic, Social, Professional & Similar

    Organizations

    Petroleum Refineries

    put approach. Rev Econ Stat 1970;52(3):262-71.

    Tan RBH, Wijaya D, Khoo HH. LCI (Life cycle inventory) analysis of fuels and electric-

    Computer System Design Services Pharmaceutical Preparation

    Manufacturing Dry Cleaning & Laundry Services Postal Service

    Electronic Computer Manufacturing Power Generation & Supply

    ity generation in Singapore. Energy 2010;35:4910-6.

    Tarancon MA, del Rio P, Callejas F. Determining the responsibility of manufacturing sectors regarding electricity consumption. The Spanish case. Energy 2011;36:46-52.

  20. Lenzen M, Pade L-L, Munksgaard J. CO2 multipliers in multi-region input-output models. Econ Syst Res 2004;16(4):391-412.

    Electronic Equipment Repair &

    Maintenance

    Primary Battery Manufacturing

    Wiedmann T, Lenzen M, Turner K, Barrett J. Examining the global environmental impact of regional consumption activities–part II: review of input-output models for the as-

    Employment Services Printing

    Hand Tool Manufacturing Real Estate Health care & Social Assistance Retail Trade

    Industrial Gas Manufacturing Services to Buildings & Dwellings

    sessment of environmental impacts embodied in trade. Ecol Econ 2007;61:15-26.

    Bailey JA, Amyotte P, Khan FI. Agricultural application of life cycle iNdeX (LInX) for effective decision making. J Clean Prod 2010;18:1703-13.

  21. Rigamonti L, Grosso M, Giugliano M. Life cycle assessment of sub-units composing a

    Insurance Agencies, Brokerages & Related Services

    Internet Service Providers & Web

    Search Portals

    Soap & Cleaning Compound Manufacturing

    Software Publishers

    MSW management system. J Clean Prod 2010;18:1652-62.

    Kennedy C, Steinberger J, Gasson B, Hansen Y, Hillman T, Havranek M, et al. Method- ology for inventorying greenhouse gas emissions from global cities. Energy Policy 2010;38:4828-37.

    Investigation & Security Services Stationary Product Manufacturing

    Baboulet O, Lenzen M. Evaluating the environmental performance of a university.

    Light Truck & Utility Vehicle Manufacturing

    Miscellaneous Electrical Equipment Manufacturing

    Surgical Appliance & Supplies Manufacturing Telecommunications

    J Clean Prod 2010;18:1134-41.

    Scipioni A, Mastrobuono M, Mazzi A, Manzardo A. Voluntary GHG management using a life cycle approach. A case study. J Clean Prod 2010;18:299-306.

  22. Carnegie Mellon University Green Design Institute. US: Economic Input-output Life Cycle Assessment (EIO-LCA); 2002[Purchaser Price Model, Available at www.eiolca.net].

    Natural Gas Distribution Travel Arrangement & Reservation

    Services

    United States Census Bureau. Annual Estimates of the Resident Population for Incor- porated Places over 50,000, Ranked by July 1, 2012 Population: April 1, 2010 to

    Non-depository Credit Intermediation & Related Activities

    Non-Residential Maintenance &

    Repair

    Truck Transportation Warehousing & Storage

    July 1, 2012. Retrieved from www.census.gov. [on May 27, 2013].

    United States Census Bureau. County totals datasets: population, population change and estimated components of population change: April 1, 2010 to July 1, 2012. Retrieved from www.census.gov. [on April 25, 2013].

    Office Furniture Manufacturing Waste Management & Remediation

    Services

    Office Supplies Excluding Paper Water, Sewage & Other Systems Abbreviation: CI, 95% confidence interval.

    United States Department of Labor Bureau of Labor Statistics. Consumer Price Index

    History Table. Retrieved from www.bls.gov. [on June 21, 2013].

    United States Census Bureau. Monthly population estimates for the United States: April 1, 2010 to July 1, 2013 (NA-EST2012-01). Retrieved from www.census.gov.

    [on July 18, 2013].

    United States Energy Information Administration. Weekly Retail Gasoline Prices. Retrieved from www.eia.gov. [on June 28, 2013].

  23. United States Energy Information Administration. Natural Gas Prices. Retrieved from

    www.eia.gov. [on June 28, 2013].

    United States Energy Information Administration. Conversion factors. Annual Energy

    Outlook; 2013 [Retrieved from www.eia.gov on June 28, 2013].

    United States Environmental Protection Agency. Primary energy production by source. Retrieved from www.epa.gov. [on December 19, 2013].

  24. United States Census Bureau. NAICS Index File; 2012[Retrieved from www.census. gov on August 5, 2013].
  25. United States Energy Information Administration. Net generation from coal. Electric Power Annual; 2012 [Retrieved from www.eia.gov on 18 June 2014].
  26. Brown LH, Buettner PG, Canyon DV, Crawford JM, Judd J. Estimating the life cycle green- house gas emssisions of Australian ambulance services. J Clean Prod 2012;37:135-41.
  27. Brown LH, Blanchard IE. Energy, emissions and emergency medical services: policy matters. Energy Policy 2012;46:585-93.
  28. Blanchard IE, Brown LH. Carbon footprinting of North American emergency medical services systems. Prehosp Emerg Care 2011;15:23-9.
  29. Su B, Huang HC, Ang BW, Zhou P. Input-output analysis of CO2 emissions embodied in trade: the effects of sector aggregation. Energy Econ 2010;32:166-75.

Leave a Reply

Your email address will not be published. Required fields are marked *