Dept. of Economics (Web page)

Karlstad University

SE-651 88 Karlstad Sweden

Phone: +46-54-700-10-00

Karlstad University Working Paper in Economics

# 2013 / 2

Time is money, but how much?

The monetary value of response time for Thai ambulance emergency services

Dr. Henrik Jaldell a, Dr. Lebnak P b, Dr. Anurak A. b, Ms. Krongkan B., Ms. Khanisthar P. b

a Department of Economics, Karlstad University

b Emergency Medical Institute Thailand, EMIT

2

Time is money, but how much?

The monetary value of response time for Thai ambulance emergency services

Dr. Henrik Jaldell Karlstad University

Department of Economics S-651 88 Karlstad

Sweden henrik.jaldell@kau.se Phone: +46547001369

Fax: +46547001799

Dr. Lebnak P., Dr. Anurak A., Ms. Krongkan B., Ms. Khanisthar P. Emergency Medical Institute Thailand, EMIT,

Bangkok, Thailand

Abstract:

The monetary values for how much ambulance emergency services are calculated for two different

time factors, response time, which is the time from when a call is received by the EMS call-taking

centre until the response team arrives at the emergency scene, and operational time, which is the

time from alarm to the accident scene and to the hospital. The study is performed in three steps.

First, marginal effects of reduced fatalities and injuries for a minute change of the time factors are

calculated using logistic regressions. Second, monetary values are chosen for fatalities and injuries;

third, the marginal effects and the monetary values are put together to find a value per minute. The

values are found to be 5.5 million Thai Baht per minute for fatality, 326,000 Baht per minute for

severe injury, and 2,100 Baht per minute for slight injury. The total value of fatality, severe injury

and slight injury for a one-minute improvement for each dispatch, summarized over one year, is 1.6

billion Thai Baht using response time. The resulting total values could be used on the benefit side in

an economic cost-benefit analysis of investments, such as new technology, which could reduce the

response and operational times.

Keywords:

Response time, cost-benefit, medicine, emergency, EMS

JEL codes:

D61, I31, R53,

Acknowledgement:

Financial support from Swedish Ministry of Foreign Affairs is acknowledged. Many thanks to Anders

Edberg, Ericsson (Thailand), without whom this project would not have been possible.

3

1. Introduction

The success of all emergency responses is dependent on the time taken to get to the place where

someone is lying ill or where a traffic accident has occurred. The faster the response the better the

outcome will be. Hence, it is reasonable to say that all efforts should be made to decrease the time

factor in the alarm chain from calling to taking the call, to dispatching, to getting ready to leave, to

driving to the injured or accident, to taking care of the injured or suppressing the fire, and to getting

the injured to hospital. On the other hand, should all efforts be made solely to decrease the time

factor? Such efforts are costly and there are other health matters that could be invested in: better

ambulances with more technical equipment, more training of the staff, better hospitals, provision of

self-help equipment etc. The economical way of dealing with this problem of the public sector is to

perform cost-benefit analyses. If benefits outweigh costs, in monetary terms, then an investment

should be made since it can be said to increase welfare in society. If costs outweigh benefits, the

investment should not be made.

The purpose of this study is to find a monetary value for the time factor of the emergency responses

in Thailand. It is not a cost-benefit analysis, since it only considers the benefit side of the time factor.

Notwithstanding, the results of the study could be used in a cost-benefit analysis. For example, if the

Thai emergency sector intends to invest in new alarm technology that could save 1 minute in

response time for all responses, how much will such an investment lead to in benefits measured in

economic welfare terms?

As noted by Blanchard et al. (2012), there are only a few studies on the relation between the

response time of emergency medical service (EMS) and the saving of lives. When it comes to cardiac

arrest, reducing ambulance response time has been shown to increase survival rate (Pons et al.

2005; Pell et al. 2001; O’Keefe et al. 2011). Gonzales et al. (2009) found increased EMS pre-hospital

time to be associated with higher mortality rates. Using fire and rescue services, which have shorter

response times than traditional ambulances for health care responses, has been found to increase

survival rate (Mattsson and Juås 1997; Jaldell 2004; Sund et al. 2011). However, there are also

studies that have concluded that there is no relation between the response time and outcome of the

patient (Blackwell et al. 2002; Blackwell et al. 2009; Pons and Markovchick 2002).

There are five motivations behind this paper. The first is that, as noted above, there is not much

research done on the effect of the response time. The second is that most of the studies mentioned

have taken up one health problem (cardiac arrest), while from a planning perspective there are of

course many more reasons for having ambulance services. Furthermore, most of the analyses have

evaluated the 8-minute response time goal for American ALS units responding to life-threatening

events, for example, by comparing the survival rate below or above the 8-minute response time

using non-continuous measures of response time. This analysis focuses instead on a continuous

measure of the response time. The third is that this study examines not only the relation between

response time and mortality, but also the effect of the illness condition for non-mortality cases. The

fourth is that the number of observations in this study is over a million, compared to hundreds or

thousands in the papers mentioned above. The fifth is that the analysis done does not stop at the

outcome of the patient, but instead takes on an economic perspective, where the purpose is to find

a monetary value for the total benefits of reducing the response time. This value could be used in a

4

cost-benefit analysis for evaluating investments in new alarm technology that would speed-up the

response time.1

To find the monetary value of the time factor for emergency responses in Thailand, the analysis is

performed in two steps. The first step is to analyze the emergency response data from the call-

taking and dispatch centre database of the Emergency Medical Institute of Thailand. The data used is

for 19 months (from March 2009 to September 2010) with 1,160,391 emergency response records

representing 73 % of all emergency response cases in Thailand during this time period. In the

statistical analysis a logistic regression analysis is used to find the relation, expressed as marginal

effects, between an independent variable and dependent variables. The dependent variables are

fatality, severe injury and slight injury. The independent variable is the response time or the

operational time, i.e. the time factor of the emergency response. Holding other independent

variables and risk factors constant, the marginal effect describes the increase or decrease in the time

factor for a one minute change and how this will affect the risk of fatality, severe injury and slight

injury.

Using results from a Thai cost-of-illness study (Thanirananon et al. 2008) the total value of fatality,

severe injury and slight injury for a one-minute improvement for each dispatch summarized over

one year is 1.6 billion Thai Baht for response time, where response time is the time from when a call

is received by the EMS call-taking centre until the response team arrives at the emergency scene. For

operational time, it is 800 million Thai Baht, where operational time is the time from when a call is

received by the EMS call-taking centre until the patient is admitted to a hospital emergency room.

The above values for a one-minute improvement to the time factor for one year are calculated using

the provinces included in the Narenthorn database. The number of emergency response cases in

these provinces represents 73 % of the total number of the emergency responses in Thailand during

the study period. Therefore, if we were to extrapolate the loss values for the whole of Thailand the

value would be 2.2 billion Thai Baht for response time and 1.1 billion million Thai Baht for

operational time. These figures represent the positive welfare effect, for one year, of reducing the

emergency responses in Thailand by one minute on average.

Assuming, for example, that an investment could be made in a new call taking and dispatch system

with a technology life of 20 years, which could decrease the response time and operational time by

one minute, the present value of the benefits of such an investment will be between 12.8 and 25.6

billion Thai Baht, assuming a social discount rate of 6 %.

Section 2 describes the Thai emergency system and section 3 contains the data used. The model and

the results are presented in sections 4 and 5, respectively. Section 6 concludes the study with a

discussion and conclusion.

1 No similar cost-benefit study has been found and there have been very few economic studies of out-of-

hospital emergency care (see Lerner et al. 2006).

5

2. Emergency System in Thailand

Currently, the emergency call number “1669” is being used as the emergency medical contact

number in Thailand. The system has been installed in each province at the main hospital or the

provincial health office. The call taker asks the caller for information and tries to understand the

symptoms or other relevant information. He/she then gives the caller some essential medical

suggestions and advice, such as first-aid, and then asks for further information about the location

and situation to be able to make a decision about the next step. A dispatcher controls the resources

by using different EMS-levels including the first response unit (FR), the basic life support unit (BLS)

and the advanced life support unit (ALS). He/she also addresses their suitability to operate at the

scene of the problem and their capacity to aid the patient. The FR-unit is able to assess and give

primary care to the emergency patient, e.g. first-aid and simple procedures. The BLS-unit has more

capability to take care of the emergency patient than the first response unit, e.g. basic medical

operation, oxygen giving and non-invasive emergency care. The ALS-unit has the capability to

provide care similar to the emergency unit in a hospital, e.g. CPR (Cardiopulmonary resuscitation)

with defibrillator, ventilation support, intravenous infusion, intravenous injection and invasive

treatments. The important role of the call taking and dispatch system is to receive the correct

information quickly, to evaluate the situation and to supply personnel, vehicles, equipment, etc,

which can support the emergency case in the best way possible and reach the location of the

incident rapidly, especially to assist an emergency patient who could be severely injured or die if the

assistance is delayed. There are 12 million emergency cases per year, 30% of which are for critical or

emergency patients, i.e. those who need the emergency services to prevent life threatening

situations. Of the total amount of emergency cases, approx. 60,000 emergency patients died

outside hospitals. If Thailand had an efficient emergency medical service, 15 – 20% of emergency

patients, or 9,000-12,000 people would be saved per year.

6

3. Data

Definition of response time and operational time

The emergency operation system can be described as having the operational flow shown in figure 1.

Figure 1: Emergency Medical Time

T0 – T1 is the time from when the person who sees or is involved in the incident makes a decision to

call the emergency number 1669 in order to request for medical assistance. This time cannot be

measured accurately because the caller cannot always accurately recall or measure the time (in

minutes) from seeing or being involved in the incident to the time of calling the emergency medical

service. T1 - T2 is the time between the caller making a phone call to the emergency services (1669)

and the call-taker answering the call, which is usually 5-10 seconds. In the case of a call taker being

unavailable, the communication supplier for the emergency operation will generally place the

emergency phone call into a queuing system; the call is connected as soon as the next free call taker

is available. T2 – T3 is the time from the call taker collecting data from the caller to when he/she

makes a decision to dispatch the appropriate emergency operation unit to the scene of the incident.

The necessary data is the location, the patient’s details, symptoms, the safety of the location, etc.

The duration might be between 15 seconds and several minutes depending on the severity and

complexity of the incident. T3 – T4 is the time from when the commander informs and dispatches

7

the emergency operation unit, until the unit vehicle leaves from its base. Normally, this will depend

upon the technology of the communication system used for transferring the entire case data to the

emergency operation unit. Also, it will depend upon the call procedure for the unit staff and the

distance between the base and their vehicle. Several emergency units are specified to move out of

the base within 1 minute after being informed of the incident, but this has not been implemented

officially, and cannot be considered as the standard service as of yet.

T4 – T5 is the time taken for the vehicle to move from the base to the incident location. T5 – T6 is

the time from arriving at the location until reaching the patient. This might differ; for example, for a

traffic accident it may take less than 15 seconds. Alternatively, if the incident is in a skyscraper in the

city centre, it will take longer (e.g. 5 minutes) to arrive at the patient’s side. T6 – T7 is the time it

takes to deliver medical care at the location, which will most likely be different from case to case.

For example, for a patient involved in a traffic accident, it will be more advantageous if he/she can

arrive at a hospital rapidly and receive medical care in the operating room as fast as possible (Scoop

and Run). On the other hand, if the patient has cardiac arrest symptoms, it will be more

advantageous if he/she can receive the necessary invasive care at the location until the situation is

stabilized, and then he/she can be transferred to the hospital (Stay and Play). T7 – T8 is the time

taken to transfer the patient to the hospital. This may differ depending on the urgency. The

decision to take the patient to the hospital will be taken by the unit leader and confirmed by the

commander, who receives the report of the emergency patient from the operation unit before

arriving at the hospital.

In this study response time and operational time are defined as:

Response Time: the response time is the time from when the call taker receives the phone call until

the operational unit arrives at the scene site. (T2-T5)

Operational time: the time from when the call taker receives the phone call to the operational unit

transfer of the patient to the hospital. (T2-T8)

The Emergency Medical Institute of Thailand (EMIT) creates the monitoring and implementation

report by extracting relevant data and information from the online-dispatch system called the

“Narenthorn Emergency Medical Database”. The local agencies report data through this system in

order to obtain financial reimbursement for the emergency medical operations they have

successfully performed. The reports in the system include basic information on the dispatch centre,

location and notification, but also time information and information about the injury. The

information consists of the time the information is received, the command time, the vehicle dispatch

time, the scene arrival time, the scene departure time, the hospital arrival time, the base returning

time, the total response time, the distance (in kilometres) and the type of operation unit.

The information on accident or emergency injury is categorized into 12 items, and for disaster into 6

items. There is also categorized information of the injury based on seriousness levels, type of

operation unit and operational staff. The reports also include information on the preliminary

operation results on scene categorized by the type of treatment and identified by the referral, for

example, death and no treatment, heart attack, onsite treatment, etc. The hospital treatment

consists of admission time, treatment duration, treatment result, referrals, continuous treatment,

death, etc.

8

The Narenthorn database has been used nationwide, except for eight provinces, and covers the

regions with about 3/4 of the population of Thailand.2 For the period studied here, 2009 – 2010,

there are 1,489,800 reports, or 73.2% of total reports, which are generated through the system.

However, there are problems with the reports from October 1, 2009 to March 31, 2010. Some

obviously contain wrong time data, for instance, a response time of over 248 minutes and an

operational time of 314 minutes3, so in total only 1,186,067 reports are used in the analysis.

Descriptive statistics

Treatment results have been categorized into three levels: slight injury, severe injury and fatality.

Slight injury means all patients who receive medical care on scene or at the hospital. Recovery is

allowed to take place at home before or after the rescue services arrive at the scene or after the

patients have received emergency care. Severe injury means patients who receive medical care, and

are admitted to a hospital, and when there is no death before or after the rescue arrives on the

scene, or after the patients receive emergency care. Fatality means patients who die before or after

the rescue services arrive at the scene, or after the patients receive emergency care, and includes

death at the hospital.

Cause of incident is divided into four groups: physical trauma, medical emergency, traffic accident

and others. Physical trauma includes a fall and collapse, fall from a height, building collapse, physical

assault, trauma from an external object, trauma from an animal, fire, electrocution, burns, bombing,

natural hazards, and hazmat. Medical emergency includes drowning, suicide and medical

emergency, while traffic accident includes motor vehicle collision. The number of dispatches for

each incident group with regard to EMS-level and treatment result is found in tables 1a- 1c. Medical

emergency is the most frequent cause of incident, followed by traffic accidents. ALS-units are more

often dispatched to medical emergencies than BLS- and FR-units, while BLS-units are more often

dispatched to traffic accidents. It can be seen that ALS-units are dispatched to a higher degree to

more serious injuries, followed by BLS-units and FR-units. In tables 2a-2b the response and

operational times are reported for different EMS-levels and treatment results. ALS-units also have

the longest response times followed by BLS-units and FR-units. However, the operational time is

similar for all three units.

2 The provinces not included are Bangkok, NongKhai, NongBualamphu, Udonthani, Kalasin, Khonkaen,

Mahasalakham and Roiet. 3 The maximum time is chosen according to mean + one standard deviation.

9

Table 1. Number of dispatches for each EMS-level and treatment results. a. Total EMS LEVEL

EMERGENCY ALS BLS FR n n n n

Medical emergency 670,313 56.5% 117,560 64.4% 139,085 53.7% 413,668 55.5% Traffic accident 358,173 30.2% 47,523 26.1% 83,237 32.1% 227,413 30.5%

Physical trauma 128,410 10.8% 13,491 7.4% 29,370 11.3% 85,549 11.5% Other 29,171 2.5% 3,845 2.1% 7,227 2.8% 18,099 2.4%

Total 1,186,067 100.0% 182,419 100.0% 258,919 100.0% 744,729 100.0%

b. Treatment results

EMERGENCY Total FATALITY SEVERE SLIGHT n n % n % n %

Medical emergency 670,622 56.5% 12,476 58.7% 180,126 62.0% 462,082 56.3%

Traffic accident 358,435 30.2% 6,915 32.6% 71,393 24.6% 247,374 30.1% Physical trauma 128,478 10.8% 1,694 8.0% 26,814 9.2% 95,392 11.6%

Other 29,207 2.5% 151 0.7% 11,971 4.1% 16,119 2.0%

Total 1,186,742 100.0% 21,236 100.0% 290,304 100.0% 820,967 100.0% FATALITY=worst of injuries, SEVERE=worst of injuries, SLIGHT=worst of injuries.

c.

EMS LEVEL Total FATALITY SEVERE SLIGHT

ALS 182,419 15.4% 14,647 69.0% 94,046 32.4% 62,994 7.7%

BLS 258,919 21.8% 2,372 11.2% 67,376 23.2% 173,196 21.1% FR 744,729 62.8% 4,205 19.8% 128,814 44.4% 584,275 71.2%

Total 1,186,067 100.0% 21,224 100.0% 290,236 100.0% 820,465 100.0%

FATALITY=If fatality was worst of injuries, SEVERE=If severe injury was worst of injuries, SLIGHT=If slight injury was worst of injuries.

Table 2. Percent of each treatment and response and operational time in minutes for each emergency group and for each EMS-level. a.

EMERGENCY FATALITY %

SEVERE %

SLIGHT %

Response time

Median

Response time

Mean

Response time Std

Operational time

Median

Operational time

Mean

Operational time Std

Medical emergency 1.9% 26.9% 68.9%

9 37.6 206.5 26 66.3 241.0

Traffic accident 1.9% 19.9% 69.0% 7 38.4 221.5 19 67.3 260.9 Physical trauma 1.3% 20.9% 74.2% 7 36.7 210.5 23 65.0 244.5

Other 0.5% 41.0% 55.2% 9 37.9 208.7 29 69.8 247.7

Total 1.8% 24.5% 69.2% 8 37.8 211.6 24 66.6 247.7

b. EMS LEVEL FATALITY

% SEVERE

% SLIGHT

% Response

time Median

Response time

Mean

Response time Std

Operational time

Median

Operational time

Mean

Operational time Std

ALS 8.0% 51.6% 34.5% 12 36.6 191.4 25 61.9 225.9 BLS 0.9% 26.0% 66.9% 9 30.2 177.3 23 61.5 224.6

FR 0.6% 17.3% 78.4% 7 40.7 226.8 24 69.5 260.2

Total 1.8% 24.5% 69.2% 8 37.8 211.6 24 66.6 247.7

10

In figure 2 we can see the relation between the response time variable and the percent of death and

severe injury for all cases and for each emergency type. The risk of fatality increases by up to a

response time of 20-25 minutes, but after 25-30 minutes the curves seem to be quite horizontal and

thus the risk of dying is no longer increasing.

Figure 2. Proportion of fatalities related to response time.

For severe injuries the relations have about the same shapes (not shown here). There is an increased

risk of a severe injury for shorter response times, but after about 30 minutes (shorter for traffic

accidents) a longer response time no longer leads to an increased risk of a severe injury.

3.2 Monetary value of emergency injury or accident

The purpose of an economic cost-benefit analysis, CBA, is to measure the welfare impacts of public

investments. If the benefits of the investment are larger than the costs, measured in monetary units,

then welfare can be increased by investing in the project. Therefore, in this analysis we need figures

in Thai Baht for saving lives and reducing injuries.

There are two main methods of finding such monetary values: the cost-of-illness (COI) method and

the willingness to pay (WTP) approach. WTP is based on the idea that people can assess the risk of

having an accident, and that they will pay for reducing or minimizing that risk (see e.g. Viscusi and

Aldy, 2003; Bellavance et al., 2009; Lindhjem et al. 2011). The monetary value is derived either from

questions asked of people (stated preference technique) or by studying people’s behaviour, e.g. how

much they pay when buying risk reducing protection or how high a wage they want for accepting a

job with a higher risk (revealed preferences).

11

When it comes to estimating the value of a statistical life, VSL, there have been only a few studies

that cover Thailand. Vassanadumrongdee and Matsuoka (2005), using surveys in Bangkok with 1,080

questionnaires (680 for the air pollution sample and 400 for the traffic accident sample), employed

the stated preference technique contingent valuation to estimate VSL in the context of air pollution

and traffic accidents. For both risk contexts they used the same reductions in risk level with

reductions of 30/1000000 and 60/1000000. The income adjusted VSL was found to be 59 million

Baht for the smaller risk reduction and 38 million Baht for the larger for air pollution, and 61 million

Baht for the smaller risk reduction and 38 million Baht for the larger for traffic accidents. Chestnut et

al. (1998) tried to find a VSL for air pollution in Bangkok. They referred to studies done in other

countries and used a benefit transfer to calculate a value of US $0.80 to $2.78 million. Gibson et al.

(2006) calculated a VSL of US $0.25 million for landmine clearance in rural Thailand using the

contingent valuation method. Miller (2000) compared the VSL of transport between different

countries, by means of benefit transfer using countries’ different GDP levels, to calculate “best

estimates” for each country. The “best estimate” for Thailand was US $0.38 million.4

The above studies only calculate values of a statistical life. However, we are also interested here in

the monetary value of severe injury and slight injury. For our monetary values results from a study

by Pichai Thanirananon et al. (2008), which employed a cost-of-illness method to calculate the cost

of traffic accidents in Thailand in 2004, has been used. The cost-of-illness method is a way to

calculate the consequences of accidents in monetary values (see e.g. Tarricone, 2006; Larg and

Moss, 2011). That is, it is the sum of emergency costs, hospital costs, productivity loss etc.

Thanirananon et al. focused on five regional hospitals which had a department for providing service

data on the injuries caused by traffic accidents. The loss value was categorized into 3 groups as

follows: 1) The human cost group (loss of productivity, quality of life, medical costs, emergency

medical service costs, long-term costs, etc) 2) The damaged property cost group (vehicle and other

properties damages). 3) The crash cost group (management expenditure of insurance companies,

police, courts, rescue services and the delay of transportation). The loss value for 2004 was also

recalculated to values for fatality of 63,317 million Baht, for severe injury 58,963 million Baht and for

slight injury 1,299 million Baht for 2011 by adjusting for inflation (increasing by 25 %).

4 Another question is whether the same value should be used for different injuries; some studies have found

different values depending on the context (e.g. Savage, 1993; Jones-Lee and Loomes, 1995; Hammitt and Liu, 2004; Carlsson et al., 2010). However, this problem has not been taken into account in this study.

12

4. The Model

We have chosen logistic regressions because of the structure with binary dependent variables. The

problem is choosing how to find a model that both best fits the data and performs well in calculating

the marginal effects of a change in response time that is true for all dispatches. For an example of

how this can be discussed, let us look at the relation between response time and deaths in traffic

accidents in figure 2. Since there seems to be no change in deaths after about 25 minutes, one

choice of model is to restrict the data to only those dispatches where the response time is less than

or equal to 25 minutes. The problem with such a model is that it will predict a much higher

proportion of deaths above 25 minutes than is reasonable according to the data. Consider figure 2

where we can see that about 5.5 % deaths is reasonable for a response time of 40 minutes.

However, a logistic regression model that is restricted to less than 25 minutes would predict this to

be about 40 %. Another suggestion is to choose something like a moving average logistic regression

model, where the first model includes only data for 1 to 5 minutes, the second from 2 to 6 minutes,

the third from 3 to 7 minutes and so forth. Predictions and marginal effects are then calculated for 3

minutes for the first model, 4 minutes for the second model and so forth. Such a model fits the data

much better, but is not very general of course since it has different parameter values for each

minute of response time. Yet another alternative is to try to include as many data points as possible.

This is used here and all response times including median time + one standard deviation are

included. All three models are shown in figures 3 (predictions of proportions of deaths) and 4

(marginal effects).

What we are after is a value for a change of one minute in response time for an average dispatch.

Here, we use the model with the median + 1 standard deviation for response time included, even if

it does not fit the data perfectly. However, choosing one of the other two models would result in

much too high a marginal effect for an average dispatch. The models thus contain response times up

to 249 minutes and operational times up to 313 minutes. Since the relation between the outcome

and the response time seems to be somewhat different, depending on the case of the emergency,

we have chosen to perform different statistical analyses for each case of emergency (traffic

accidents, medical emergency, physical trauma and others).

0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0 5 10 15 20 25 30

Pro

po

rtio

n d

ead

Response time, min

Pred value moving average

Pred value response time <=249min

Pred value response time <=25 min

Figure 3. Relation between response time and predicted proportion of deaths using different models.

13

-0.003

-0.002

-0.001

0

0.001

0.002

0.003

0.004

0.005

0.006

0.007

0 5 10 15 20 25 30

Pro

po

rtio

n d

ead

Response time, min

Marg eff moving average

Marg eff response time <=249 min

Marg eff response time <=25 min

Figure 4. Relation between response time and marginal effect for proportion of deaths using different

models.

14

5. Results

Since the dependent variables have been set to be binary, logit regression analyses have been used

to find out the relation between the independent variables, response time and operational time, and

the dependent variable. The parameter estimates for the independent variables are recalculated

into marginal effects, which show how much the risk of fatality, severe injury and slight injury

changes when the time variable is changed by one minute.

The analyses have thus proceeded in three steps. First, logistic regression models have been used to

find parameter estimates for how the time variables affect the three injury types (equation 1).

Equation 1 has been estimated for each injury and emergency type, and for response and

operational time respectively, that is 3*4*2=24 models have been estimated.

*

*( ) Prob( 1)

1

TIME

TIME

eE Y Y

e

(1)

Second, since the model is nonlinear the parameter estimates have been recalculated into marginal

effects (equation 2). The marginal effects are evaluated at the median response and operational

time.5

*

* 2

( )Marginal effect=

(1 )

TIME

TIME

E Y e

TIME e

(2)

The marginal effects for response time are presented in table 3. They are higher for severe injury

than for fatalities, meaning that a marginal decrease of response time leads to more people saved

from severe injury than from fatality. For fatality the marginal effect is highest for traffic accidents,

while for severe injury it is highest for others followed by medical emergency. For slight injuries the

marginal effects are negative and will therefore not be used in the next step. For operational time

(not showed here) the marginal effects are lower than for response time, indicating that there is a

decreasing marginal value of time, since operational time is longer than response time.

Third, the marginal effects have been recalculated into number of persons affected by a minute

change in response and operational time in one year (equation 3), as presented in table 4 and 5. If

the marginal effect is not statistically significant or negative, the value is set to zero.

*

* 2

( )Marginal effect in one year= * *

(1 )

TIME

TIME

E Y en n

TIME e

, (3)

where n is equal to total number of responses in one year for each emergency type. A one-minute

change would save most people from fatality when it comes to traffic accidents. For severe injuries a

one-minute change would save most in the treatment group others, followed by medical

emergency.

Fourth, the monetary values have been summed up in Thai baht, ฿, for one year, for each

emergency type and totally for all emergency responses. Using the monetary values of lives and

5 Normally marginal effects are evaluated at the sample means of the data or the sample averages of the

individual marginal effects are used (Greene 2008). However, since the median in the sample used here better describes the typical response and operation time than the mean value does, the median has been used here.

15

injuries, we can calculate a total value per EMS type. The results are shown in table 6. For both

response time and operational time the most important treatment type is medical emergency,

followed by traffic accident. The values for operational time are lower than the values for response

time, reflecting the decreasing marginal value of time. However, the ratio between response and

operational time differs for the different emergency types. The relative difference is smallest for

traffic accident and largest for others, and about 1/2 for the total emergencies. Different ambulance

types have different marginal benefit values per minute. For response time, ALS has a value of 1130

Baht per minute, BLS a value of 644 baht per minute and FR a value of 445 baht per minute.

Table 3. Marginal effects and P(.) >0 results for response time evaluated at median response time (=8 min).

Injury type / emergency type

Physical Trauma Medical Emergency

Traffic Accident Others

Fatality 0.0001473

(0.000) 0.0001912

(0.000) 0.0002861

(0.000) 0.0000287

(0.309)

Severe injury 0.0027129

(0.000) 0.0040699

(0.000) 0.0017932

(0.000) 0.0047531

(0.000)

Slight injury -0.0004476

(0.000) -0.0002977

(0.000) -0.001437

(0.000) -0.0003409

(0.000)

Table 4. Deaths and injuries saved per year calculated given marginal effect per minute for response time.

Injury type / emergency type

Physical Trauma Medical Emergency

Traffic Accident

Others

Fatality 11.9 15.5 23.2 2.2 Severe injury 220.0 330.0 145.4 398.3 Slight injury 0 0 0 0 Number of dispatches 81101 423356 226215 18424

Table 5. Deaths and injuries saved per year calculated given marginal effect per minute for operational time.

Injury type / emergency type Physical Trauma Medical

Emergency

Traffic Accident Others

Fatality 8.8 8.7 17.0 -

Severe injury 88.5 109.8 51.4 136.5

Sligth injury 0 0 0 0 Number of dispatches 81101 423356 226215 18424

Table 6. Monetary value per minute and year. Baht/Year/Minute/

emergency type

Physical Trauma Medical

Emergency

Traffic Accident Others Total

Response time

(at median 8 minutes)

฿ 135,401,000 ฿ 987,186,000 ฿ 484,352,000 ฿ 27,349,000 ฿ 1,634,289,000

Operational time

(at median 24 minutes)

฿ 76,177,000 ฿ 427,974,000 ฿ 304,957,000 ฿ 9,370,000 ฿ 818,477,000

The loss values for a one-minute improvement in the time factor for one year are calculated using

the provinces in the Narenthorn database. Eight provinces, including Bangkok, are not included in

the Narenthorn database. The number of emergency response cases in these provinces represents

16

26.8% of the total number of the emergency responses in Thailand during the period considered

here. Therefore, if we were to extrapolate the loss values for the whole of Thailand we should,

therefore, increase the total loss value by dividing the study result with 73.2%.

The result of such an extrapolation for a response time is 2,232,600,000 Thai Baht and for an

operational time 1,118,100,000 Thai Baht. These figures represent the positive welfare effect, for

one year, of reducing the emergency responses in Thailand by one minute on average.

17

6. Discussion and conclusion

This study shows that using a logistic regression analysis makes it possible to find a correlation

between response time and the severeness of injury. The correlation indicates that a faster response

time results in fewer fatalities and milder severeness of injury. Furthermore, the time factor is most

important for medical emergency, followed by traffic accidents and physical trauma. The results also

show that the more advanced the ambulance that is used the more important the response time is.

For operational time the correlation has the same sign, but it is not as strong as for response time,

which seems reasonable since there should be a decreasing marginal utility of time.

One limitation of the study is that the emergency response data cannot categorize permanent

disability as a final outcome; thus, the additional loss value of disability is excluded in the analysis,

and the loss value for those cases is covered under the category severe injury.

The planned investment thought of here is a better alarm system which could reduce the time from

accident or injury to dispatch of ambulance, and result in a one-minute decrease in response time. In

comparison, a study in Canada showed that the introduction of base paging reduced the call-

response interval by 30 seconds (Jermyn 2000). Considering operational time, Spaite et al. (1993)

listed several observed problems on-scene, for example with communication, equipment and

uncooperative patients. Most of the time was concerned primarily with logistics and not with

medical care, and operation problems occurred in more than 40 % of the dispatches. Another way to

decrease time is to enforce a single alarm number in Thailand (as in the EU, 112, or North America,

911) instead of the different numbers to police, fire and rescue services and emergency response,

together with dialling directly to hospitals for ambulances. Thus, there seems to be possibilities for

increased effectiveness. However, high speed driving could perhaps be the solution to faster

response time in rural areas (Petzäll et al. 2011), but probably not in populated areas; and using

lights and sirens when driving ambulances has both pros and cons such as high risk of crashes

(Lemonick, 2009; see also Salvucci et al., 2004).

Assume that an investment could be made, one which could decrease the response time and

operational time by one minute: for example, a new call taking and dispatch system with a

technology life of 20 years. Using the results of this study, the present value of the benefits of such

an investment is between 12.8 and 25.6 billion Thai Baht, assuming a social interest rate of 6 %.

18

References

Bellavance, F., Dionne G., Lebeau M. (2009) The value of a statistical life: A meta-analysis with a

mixed effects regression model, Journal of Health Economics, 28 (2) 444–464.

Blackwell, T.H, Kaufman, J.S. (2002) Response time effectiveness: comparison of response time and

survival in an urban emergency medical services system, Academic Emergency Medicine, 9(4) 288-

295.

Blackwell, T.H., Kline, J.A., Willis, J.J., Hicks, G.M. (2009) Lack of association between prehospital

response times and patient outcomes. Prehospital Emergency Care, 13(4), 444-50.

Blanchard I.E., Doig C.J., Hagel B.E., Anton A.R., Zygun D.A., Kortbeek J.B., Powell D.G., Williamson

T.S., Fick G.H., Innes G.D. (2012) Emergency medical services response time and mortality in an

urban setting, Prehospital Emergency Care, 16(1):142-51.

Carlsson, F., Daruvala, D., Jaldell, H. (2010) Value of statistical life and cause of accident: a choice

experiment, Risk Analysis, 30(6), 975-986.

Chestnut, L., Ostro B., Vichit-Vadakan N., Smith K. (1998). Final report health effects of particulate

matter air pollution in Bangkok. A report to the pollution control department, Thailand.

Gibson, J., Barns S., Cameron M., Lim S., Scrimgeour F., Tressler J. (2006) The value of statistical life

and the economics of landmine clearance in developing countries, World Development, 35(3), 512–

531

Gonzales, R.P., Cummings. G.R., Phelan, H.A., Muleker, M.S., Rodning, C.B (2009) Does increased

emergency medical services prehospital time affect patient mortality in rural motor vehicle crashes?

A statewide analysis, American Journal of Surgery, 197, 30-34.

Hammitt, J.K., Liu J-T. (2004) Effects of disease type and latency on the value of mortality risk.

Journal of Risk and Uncertainty, 28:73–95.

Jaldell, H. (2004) Tidsfaktorns betydelse vid räddningsinsatser, Swedish Rescue Services Agency, P22-

499. (English translated version: The importance of the time factor in fire and rescue service

operations in Sweden – an update of a socio-economic study, manuscript, Karlstad University)

Jermyn, B.D. (2000) Reduction of the call-response interval with ambulance base paging, Prehospital

Emergency Care, 4(4), 318-321.

Jones-Lee M., Loomes G. (1995) Scale and context effects in the valuation of transport safety.

Journal of Risk and Uncertainty, 11:183–203.

Lemonick, D.M. (2009) Controversies in prehospital care, American Journal of Clinical Medicine, 6(1),

5-17.

Larg, A., Moss, J. R. (2011) Cost-of-illness Studies: a guide to critical evaluation, PharmacoEconomics,

29(8), 653-671.

19

Lerner, E.B, Maio, R.F., Garrison, H.G., Spaite, D.W., Nichol, G. (2006) Economic value of out-of-

hospital emergency care: a structured literature review, Annals of Emergency Medicine, 47(6)515-

524.

Lindhjem, H., Navrud, S., Braathen, N.A., Biausque, V. (2011) Valuing mortality risk reductions from

environmental, transport, and health policies: a global meta-analysis of stated preference studies,

Risk Analysis, 31 (9), 1381–1407.

Mattsson, B., Juås B. (1997) The importance of the time factor in fire and rescue service operations

in Sweden, Accident Analysis and Prevention, 29 (6), 849-857.

Miller T.R. (2000) Variations between countries in values of statistical life, Journal of Transport

Economics and Policy, 34, 169-188.

O’Keefe, C., Nicholl, J., Turner, J., Goodacre, S. (2011) Role of ambulance response times in the

survival of patients with out-of-hospital cardiac arrest, Emergency Medical Journal, 28, 703-706.

Pell, J.P., Sirel, J.M., Marsden, A.K. (2001) Effect of reducing ambulance response times on deaths

from out of hospital cardiac arrest: cohort study, British Medical Journal, 322 (7299), 1385-1388

Petzäll, K., Petzäll, J., Jansson, J., Nordström, G. (2011) Time saved with high speed driving of

ambulances, Accident Analysis and Prevention, 43, 818-822.

Pichai Thanirananon, Chartbunchachai, Plajpalkhumthrapy, Waugh, Yxdplthnabriburn,

Ladensrisakda, Wathnawang, Phiphathntangchim (2008) National study of accidents, Faculty of

Engineering, Prince of Songkla University/ Department of Highways, Ministry of Transport.

Pons, P.T., Haukoos, J.S., Bludworth, W., Cribley, T., Pons, K.A, Markovchick, V.J. (2005) Paramedic

response time: does it affect patient survival? Academic Emergency Medicine, 12 (7), 594-600.

Pons, P.T., Markovchick, V.J. (2002) Eight minutes or less: Does the ambulance response time

guideline impact trauma patient outcome, Journal of Emergency Medicine, 23(1), 43-48.

Salvucci, A., Kuehl, A., Clawson, J. (2004) The response time myth – Does time matter in responding

emergencies?, Topics in Emergency Medicine, 26(2), 86-92.

Savage I. (1993) An empirical investigation into the effect of psychological perceptions on the

willingness-to-pay to reduce risk, Journal of Risk and Uncertainty, 6:75–90.

Spaite, D.W., Valenzuela, T.D., Meislin, H.W., Criss, E.A., Hinsberg, P. (1993) Prospective validation of

new model for evaluating emergency medical services systems by in-field observation of specific

time intervals in prehospital care, Annals of Emergency Medicine, 22(4), 638-645.

Sund, B., Svensson, L., Rosenqvist, M., Hollenberg, J. (2011) Favourable cost-benefit in an early

defibrillation programme using dual dispatch of ambulance and fire services in out-of-hospital

cardiac arrest, European Journal of Health Economics, 13(6), 811-8.

Tarricone, R. (2006) Cost-of-illness analysis. What room in health economics?, Health Policy 77, 51–

63.

20

Vassanadumrongdee, S., Matsuoka, S. (2005) Risk perceptions and value of a statistical life for air

pollution and traffic accidents: evidence from Bangkok, Thailand, Journal of Risk and Uncertainty, 30

(3), 261-287.

Viscusi, W.K., Aldy, J.E. (2003) The value of a statistical life: a critical review of market estimates

Throughout the World, Journal of Risk and Uncertainty, 27(1), 5-76.