Erasmus School of Economics, Health Economics

The cost of motor vehicle accidents caused by obstructive sleep apnea syndrome in the Netherlands

Erasmus School of Economics, Health Economics

R.M.A.E. Vroegop

Index

1. Introduction 1

2. Methods and data

2.1 Overview 3

2.2 Cost of motor vehicle accidents in The Netherlands 3

2.3 Prevalence of OSAS 5

2.4 Increased motor vehicle accident risk for OSAS patients 8

2.5 Method for calculating the proportion of preventable accidents 10

3. Results

3.1 Collected data 12

3.2 The proportion of preventable accidents 12

3.3 Preventable MVA cost 13

3.4 Legislation 13

4. Discussion 15

5. Conclusion 17

6. References

1. Introduction

Sleep apnea syndrome is characterised by an abnormal high level of respiratory arrests during sleep. There are two different pathways that will result in sleep apnea syndrome. With Central Sleep Apnea Syndrome (CSAS) the brain fails to give enough stimulation to the respiratory muscles responsible for the breathing motion. The other pathway is Obstructive Sleep Apnea Syndrome (OSAS), in this case the soft pharynx tissue and tongue muscles relax too much during sleep which causes an obstruction of the upper airway.

The American Academy of Sleep Medicine task force composed worldwide-accepted OSAS criteria.[1] An apnea is defined as a respiratory arrest for at least 10 consecutive seconds. To express the severity of apnea the Apnea Hypopnea Index is used (AHI). The arrests must occur at least five times per sleeping hour (AHI>5) to be qualified as apnea syndrome. Besides this patients must also experience extreme sleepiness during daytime or fulfil at least two of the 5 following criteria: concentration problems, daytime fatigue, unsatisfying sleep, multiple awakenings during sleep, choking or gasping during sleep.

A polysomnography allows CSAS to be distinguished from OSAS. This sleep study collects data on activity of the heart, brain and respiratory muscles. In case of CSAS the respiratory muscles are not activated during an apnea.

A respiratory arrest causes a reduction of the oxygen saturation of the blood. This hypoxemia results in the activation of the sympathetic nervous system. This system, which is part of the autonomic nervous system, recognizes this as a stress situation and will respond accordingly. This causes the arousals that patients consciouslyorunconsciously experience. Due to these arousals patients could show symptoms of extreme sleepiness and irritable or depressive feelings. Other common signs are extreme perspiration and heavy snoring during sleep.

These symptoms can have a major impact not only on the lives of patients but also their partners. The sometimes vague and unrecognised health issues that sleep apnea patients experience result in a high number of undiagnosed patients. On an individual level this even leads to a two-fold higher risk of becoming work disabled as the study of Sjösten et al. showed.[2]

Once sleep apnea is diagnosed patients are encouraged to reduce preventable risk factors. If necessary patients can use a mandibular repositioning device which prevents the tongue of retracting. Another treatment option is CPAP (Continuous Positive Airway Pressure), this device will generate a positive airway pressure applied through a mask during sleep. This will prevent respiratory arrests. When anatomical abnormalities form the basis of the problem surgery is a good treatment option.

This study will focus on OSAS because this is the most common form of sleep apnea. There are multiple risk factors for developing OSAS. Because males have a different anatomical design of the upper airway area they are more prone to develop OSAS. Also anatomical abnormalities in the upper airway area are an important risk factor. Intoxications such as smoking, alcohol consumption and the use of prescription drugs are also positively correlated with the development of OSAS. But the most important risk factor for developing OSAS is obesity. The percentage of obese people in developed countries is rapidly increasing. This will lead to an even bigger group of OSAS patients in the near future.

Several studies have shown that apnea patients have a much higher risk of being involved in motor vehicle accidents. Combined with the fact that most patients are still undiagnosed and therefore not treated this could have a large impact on the cost related to motor vehicle accidents.

In this study I will examine the cost of motor vehicle accidents in the Netherlands in which apnea patients are involved. I will try to isolate the preventable cost that are caused by undiagnosed OSAS patients. These results can give us information about the need for more active detection of OSAS in the population.

2. Methods and data

2.1 Overview

The incorporation of multiple variables is necessary in order to examine the motor vehicle accident (MVA) cost that are caused by undiagnosed OSAS. The variables that will be used are the total cost of traffic accidents in the Netherlands, the prevalence of OSAS in the population and the extra risk OSAS patients have of being involved in traffic accidents. The problem with these three variables is that there is an amount of uncertainty regarding the exact numbers.

I will therefor first explain how and why I selected certain information. Secondly, the method of merging these variables into one formula will be discussed.

2.2 Cost of motor vehicle accidents in the Netherlands

SWOV (Institute for road safety research) and DVS (Office for traffic and shipping) have conducted a thorough research together in this area called ‘Cost of traffic accidents’.[3] SWOV is an independent research company which is specialised in researching road safety issues commissioned by the government or companies from the private sector. The DVS is part of the Dutch Ministry of Infrastructure and Environment. The research was conducted between 1997 and 2003 and followed the guidelines of the COST-313 report[4] that was commissioned by the EU. It was intended to supply crucial information about the cost of traffic accidents in the Netherlands for government policy, evaluation and international comparison.

The research tried to capture both registered and unregistered accidents in order to get realistic results. The calculations were based on six main cost drivers:

-  Medical cost: Based on statistics about the average number of days victims are hospitalised. This was multiplied by the average cost per day of hospital care. And also ambulance cost were taken into consideration.

-  Production loss: The added value that victims otherwise could have produced in the future when they wouldn’t have had an accident.

All future losses caused by hospitalisation and disability are discounted for the present year. Also consumption loss in case of deaths is being accounted for.

-  Loss in quality of life: These calculations are based on the willingness to pay in order to reduce the risk of being involved in a MVA. Corrected for consumption loss, because this is already included in the production loss.

-  Material damage: Based on insurance claims, deductibles paid by victims and estimated non registered damages.

-  Case handling cost: Based on statistical information and insurance data.

-  Traffic jam cost: Cost such as lost production time due to MVA related traffic jams.

The table below shows the total cost of all six cost drivers in 2003.

Table 1. Cost of MVA in the Netherlands in 2003 in million euros

Cost driver / Cost
Medical cost / 232
Production loss / 1,294
Loss in quality of life / 5,549
Material damage / 3,866
Case handling cost / 1,262
Traffic jam cost / 125
Total Cost / 12,327

OSAS usually occurs between the age of 30 and 65 years old. Therefore most research regarding prevalence is also focussed on this particular group, as the next chapter will show.

This study will also focus on the age group of 30 to 65, which requires specifying the cost shown in table 1. The Dutch institute for road safety research (SWOV) is used to obtain the necessary data.[5] The SWOV database provides information about multiple variables of accidents. In this case the distinction between the first and second driver is important.

The person that caused the accident is marked as the first driver. In table 2 the number of accidents caused by first drivers are mentioned for each relevant age category.

Table 2. Number of motor vehicle accidents caused per age category for 2003

Age group / Number of MVA’s
30-39 / 6,240
40-49 / 4,386
50-59 / 3,248
60-64 / 904
Total / 14,778

The total number of MVA’s in 2003 was 31,635 for all age categories. So the age group of 30 to 65 was responsible for 47% of the total accidents.

In order to calculate the total MVA cost for this age group I have to assume that the cost are uniformly distributed over all age categories.

Table 1 showed that for 2003 the total MVA cost were € 12,327 million. The age group of 30 to 65 accounts for 47% of this amount which leads to € 5,793.69 million MVA cost.

2.3 Prevalence of OSAS

Regarding the prevalence of OSAS there is a lot of uncertainty. Because of the sometimes vague complaints patients have and the unawareness of doctors about OSAS a lot of patients are still undiagnosed. This causes the lack of information about the exact prevalence of OSAS. But most respected studies in this area adopt the prevalence numbers posed by Young et al.[6] This study showed that 4% of the men and 2% of the women in the age group of 30 to 65 suffer from OSAS in the United States.

These percentages are also used for the Netherlands in the Dutch OSAS treatment guidelines set by the association for doctors treating lung diseases and tuberculosis (NLVT).[7]

Regarding the number of undiagnosed OSAS patients in the Netherlands De Vries[8], among others, referred to Young’s study.

Young et al. also conducted a study to determine the percentage of OSAS patients that is undiagnosed.[9] She found that 82% of male OSAS patients with an AHI>5 were undiagnosed. For females with AHI>5 the percentage of undiagnosed patients is 93%.

The next step is to apply these prevalence findings to the situation for the Netherlands. We will focus on 2003 data because this is also the last year for which reliable MVA cost information is available. The data are obtained from the Dutch SWOV database (Institute for road safety research).[10] It must be noted that only data on the number of car driving licenses were available.

Table 3. Number of males between 30-65 years old with valid car driving license in 2003 in the Netherlands.

Age group / Number of males with a valid car driving licence in 2003
30-39 / 1,226,994
40-49 / 1,165,383
50-59 / 1,016,793
60-64 / 386,056
Total / 3,795,226

Table 4. Number of females between 30-65 years old with valid car driving license in 2003 in the Netherlands.

Age group / Number of females with a valid car driving licence in 2003
30-39 / 1,125,278
40-49 / 1,046,851
50-59 / 868,058
60-64 / 267,467
Total / 3,307,654

Tables 5 and 6 show what happens when the prevalence and undiagnosed percentages found by Young et al. are applied to the number of males and females (aged 30 to 65) with a valid driving licence in 2003.

Table 5. The number of undiagnosed male OSAS patients (aged 30-65) with a driving license in 2003

Number of males with a valid car driving licence / 3,795,226
4% OSAS prevalence.
Number of male OSAS patients / 4% x 3,795,226 = 151,809
82% Undiagnosed OSAS patients.
Number of undiagnosed male OSAS patients / 82% x 151,809 = 124,483

Table 6. The number of undiagnosed female OSAS patients (30-65) with a driving license in 2003

Number of females with a valid car driving licence / 3,307,654
2% OSAS prevalence.
Number of female OSAS patients / 2% x 3,307,654 = 66,153
93% Undiagnosed OSAS patients.
Number of undiagnosed female OSAS patients / 93% x 66,153 = 61,522

The total number of people in the Netherlands (aged 30-65) with a valid driving license is: 3,795,226 males + 3,307,654 females = 7,102,880.

Table 5 shows that 124,483 males with a valid driving license (aged 30-65) are expected to suffer from undiagnosed OSAS. In table 6 is calculated that for females this number is around 61,522. The total number of undiagnosed OSAS patients in the age group of 30 to 65 in the Netherlands is: 124,483 males + 61,522 females = 186,005

The proportion of undiagnosed OSAS patients in the group of 30 to 65 years old with a valid driving license in 2003 is:

1860057102880= 0.026

2.4 Increased motor vehicle accident risk for OSAS patients

Many studies have extensively researched this specific topic. Rodenstein reviewed a large number of these studies and found out that most studies indicated a three-fold higher MVA risk for OSAS patients.[11] Traffic densities and cultural backgrounds don’t seem to have an influence on these results.

If patients are treated with CPAP and the compliance is good, the MVA risk drops back to that of a healthy individual.

If we take a closer look at these studies two stand out because of their large sample size.

Young et al. did an epidemiologic study with 913 people, which consisted of people with and without apnea.[12] The participants had to have a polysomnographic recording that could indicate the apnea rate, a valid drivers license and had to travel at least 1600 km’s by car every year. Other variables that were taken into consideration were gender, alcohol consumption, age and education. They studied a five year period of the subjects by linking them to the transportation accident database in Wisconsin. The results showed a three-fold increased MVA risk for males with indicated sleep apnea (AHI>5). For both men and women with an AHI higher than 15 the MVA risk increased to 7 times compared to healthy people. It is strange that only men with AHI>5 have an increased risk. The researchers argue that this could be caused by the different perception of sleepiness between genders. It could also be that women act differently when feeling sleepy.