Proposed Research Plan on Vision Standard

DEPARTMENT OF TRANSPORTATION

Federal Highway Administration

49 CFR Part 391

[FHWA Docket No. MC-96-4]

Proposed Research Plan on Vision Standard

AGENCY: Federal Highway Administration (FHWA), Department of

Transportation.

ACTION: Notice; request for comments.

-----------------------------------------------------------------------

SUMMARY: The FHWA is requesting comments on a proposed research plan to

explore performance-based alternatives to the existing vision standard

for drivers of commercial motor vehicles (CMV). The findings of this

research effort may result in the modification of that standard. The

FHWA seeks comments on all aspects of the research plan, including its

scientific merit, likelihood of achieving its objective, methodological

validity, consideration of all relevant research, and other practical

issues.

The FHWA is also announcing a public hearing to obtain comments on

the proposed research plan. The hearing is designed to obtain public

input on the proposed research plan, not to determine the status of

individual drivers or participants in the vision waiver program. At the

hearing, the FHWA does not intend to discuss the status, results, or

recommendations that might result from the vision waiver program.

A review of scientific literature relevant to the vision standard

and the proposed research plan have been placed in FHWA Docket MC-96-4.

In addition, both documents are accessible electronically through the

Federal Highway Administration's World Wide Web (WWW) site.

DATES: The comment period will remain open until further notification

in the Federal Register. The public hearing will be held on August 9,

1996, at the Chicago O'Hare Marriot, 8535 West Higgins Road, Chicago,

IL, 60631, (312) 693-4444.

ADDRESSES: Submit written, signed comments to FHWA Docket MC-96-4, Room

4232, HCC-10, Office of the Chief Counsel, Federal Highway

Administration, 400 Seventh Street., SW., Washington, DC 20590.

The literature review and proposed research plan are on the Federal

Highway Administration's World Wide Web site (http://

cti1.volpe.dot.gov/ohim/whtnewhd). Users with questions about the

operation of the WWW site should call the FHWA Computer Help Desk at

(202) 366-1120.

FOR FURTHER INFORMATION CONTACT: Mr. Charles Rombro, Federal Highway

Administration, Office of Motor Carriers, 400 Seventh Street SW., room

3104, Washington, DC 20590, telephone (202) 366-5615. Office hours are

from 7:45 a.m. to 4:15 p.m., e.t., Monday through Friday, except

Federal holidays.

SUPPLEMENTARY INFORMATION: The FHWA is authorized by statute to

establish minimum physical qualification requirements for drivers of

commercial motor vehicles. 49 USC 31502.

The Congress provided the FHWA with complementary regulatory

authority with the enactment of the Motor Carrier Safety Act of 1984,

codified in substantial part at 49 U.S.C. 31101-31162. This Act

directed the Secretary to establish minimum safety standards to ensure,

inter alia, that ``the physical condition of operators of commercial

motor vehicles is adequate to enable them to operate such vehicles

safely * * *.'' 49 U.S.C. 31136(a)(3).

The physical qualification regulations for CMV drivers in

interstate commerce are found at 49 CFR 391.41. The qualification

standards cover 13 areas which directly relate to the driving function.

All but four of the standards adopted by the FHWA permit the individual

determination of a driver's qualification. A person's qualification to

drive is determined by a medical examiner who is knowledgeable about

the on-the-job functions performed by a commercial driver and whether a

particular condition would interfere with the driver's ability to

operate a CMV safely. In the case of vision, hearing, insulin-using

diabetes and epilepsy, the current standards are absolute, providing no

discretion to the medical examiner.

The current vision standard specifies that drivers must meet the

following three conditions:

1. distant visual acuity in each eye of at least 20/40, and distant

binocular acuity of at least 20/40 in both eyes; and

2. field of vision of at least 70 degrees in the horizontal

meridian of each eye; and

3. the ability to recognize the colors of traffic signals.

In order to improve protection for the public and provide for

individual determinations of fitness to drive wherever possible, the

Agency is interested in developing performance-based standards. In a

Federal Register notice on the vision waiver program published on

November 17, 1994 (59 FR 59386), the FHWA announced its intention to

initiate a research plan to ``develop parameters for performance-based

visual standards for all commercial drivers.'' 59 FR at 59389. The

research plan outlined in this notice is designed to move the Agency

towards a performance-based vision standard. This standard would

incorporate the measurement of those visual capabilities deemed

necessary for the safe operation of commercial vehicles. The research

discussed below is designed to relate specific visual functions to

specific driving tasks, such as the ability to stay in a lane. The

standards would still be prescriptive in that they would establish a

minimum score which individuals would be required to meet to be allowed

to drive; however, the scoring scheme would be based on detailed

research on the visual attributes required to safely operate a CMV.

Research Plan

The FHWA has developed a proposed research plan, an outline of

which is provided below.

Background

The FHWA's review of the existing literature on vision and driving

research led the FHWA to the following conclusions:

1. The current testing standard lacks criterion, or predictive,

validity; that is, it is not clear that central visual acuity by itself

is a good predictor of safe driving. This detracts from the perceived

fairness of the standard. The principal shortcoming of the current

standard is the emphasis on central visual acuity, which is a measure

of how well an individual can discern static images in the center of

vision under conditions of high luminance. Since many driving

situations involve dynamic images under low luminance, other visual

capacities may be equally

[[Page 28548]]

or more important than central visual acuity.

2. Improving the criterion validity of the vision standard would

most likely require testing a broader array of visual capabilities than

those included in the current standard.

3. There is no assurance that a standard based on a better

understanding of the relationship between different visual capabilities

and driving would result in a significant, measurable improvement in

safety, in part because vision is a contributor to only a small number

of crashes. CMV drivers comprise a small proportion of drivers, and are

represented in a small proportion of crashes, not all of which are

caused by failures of visual performance as currently measured.

However, the weak observed causal relationship between vision and

crashes may be a shortcoming with the current measurement of vision. It

is therefore possible that the measurement of other visual functions

could reveal a more significant and direct connection between vision

and driving ability.

The goals of the proposed research are threefold:

1. Establish a list of visual performance parameters that appear to

hold promise as a basis for a new standard, and design or adapt

performance tests to measure these capabilities.

2. Evaluate the predictive validity of these tests.

3. Based on the results of these tests, establish a trial vision

standard, and test that standard to evaluate its validity.

Outline of the Research Plan

The research contains both a long-term and a short-term track. The

Agency may decide to conduct either of the two tracks individually,

both tracks simultaneously, both tracks consecutively, or neither

track. The emphasis of the short-term plan is to build on existing

knowledge to develop an improved vision test, with the goal of adding

two or three existing vision tests to the battery currently tested.

This track does not call for significant new research, but rather seeks

to take advantage of already completed work. This track could result in

the development of a two-tier standard, with an expanded battery of

tests in the first (screening) tier, and various administrative

measures proposed for the drivers who do not pass this first tier.

Administrative measures may include provisional or restricted licenses,

waivers, or in-use monitoring of drivers.

The long-term plan consists of new research and analysis, which may

lead to the development of a new standard.

Two phases of the research are already complete. The short-term

research consists of six or seven additional stages, as explained

below. The long-term research would consist of four additional phases.

The phases of the research are described below.

Research Completed or Underway

1. Development of the research plan. This phase is complete, and

the work described below is the output of the planning effort. This

phase describes the proposed approaches in some detail; certain

elements are necessarily unspecified. For example, the choice of

specific visual function tests cannot be made before further research

and analysis are complete.

2. Design of testing strategy. This phase involves selecting and

developing the form of the candidate tests, as well as the measures

that the tests must predict. This phase is currently underway.

Selecting the candidate tests includes a general selection of

visual functions to be tested, an inventory of the tests already

available, and identification of new tests to be developed and

validated. Tests should have broad acceptance and stable underlying

population norms. The protocols for testing should be developed and

accepted by researchers and testers, and results must be stable

regardless of who administers the test. Acceptable population norms are

necessary if a test is to be used to classify individuals based on

``normal'' results in the population of CMV drivers.

While some tests, such as the Snellen Letter test of visual acuity,

are broadly accepted and have stable and well understood population

norms, other tests of potentially important visual functions do not

meet these standards. These latter tests would have to be evaluated.

The evaluation would consider how important the visual function is in

the driving task, the extent to which the test results are stable and

reliable, and how readily the tests can be developed for broad usage.

Our review of the literature has led us to focus on the following

visual functions as most relevant to the driving task: static acuity,

contrast sensitivity, dynamic acuity, working field of view, dark

focus, low contrast acuity, glare sensitivity, and vection.

The FHWA, with the assistance of a contractor, is in the process of

identifying the behaviors that the tests must predict--the measures of

effectiveness. This will be followed by the systematic development of

measures to be validated by the visual performance tests.

Theoretically, we would prefer to be able to relate a driver's

performance on vision tests to an actual driving record, especially the

driver's accident record. Because of the infrequency of accidents,

however, we would need to test a very large number of drivers over many

years to obtain reliable results. In addition, a ``clean'' experiment

would require that we allow drivers who we suspect may be deficient in

some key visual function to operate CMVs on the road. Allowing

potentially hazardous drivers to operate CMVs poses obvious safety

problems, and contravenes the FHWA's mandate to protect highway safety.

As a proxy for accidents, we are developing a list of candidate

visually- related driving behaviors. This is a reasonable proxy because

it is the driver's behavior that connects visual deficiencies and

accidents. An example might include the time a driver takes to initiate

a braking maneuver. Behaviors will be selected for further testing

based on their likely validity and practicality. Since we will not be

measuring accidents, it is especially important that the measures are

closely related to driving performance.

After choosing behavioral measures, the Agency will develop test

procedures and protocols.

Proposed Short-Term Research

3. Define criteria for selection of vision tests. The likely

criteria will include: test availability with little or no

modification, scientific reliability, construct validity, practicality

of use in a testing environment, and acceptability to researchers and

testers. While other criteria are possible, the FHWA anticipates that

the factors listed above will be used to screen tests for their

suitability for further research. Much of the work required to define

the criteria has already been completed under tasks 1 and 2.

4. Select candidate tests. The researchers would select 3 to 5

candidate tests for further research. The tests would have to meet the

criteria identified above. The researchers would determine which tests

meet these criteria through a survey of the scientific community and

other interest groups, and through the literature review conducted in

task 1.

5. Design demonstration/evaluation project. This task consists of

specifying the details of the testing procedures. The researchers would

select a site for the tests, choose criteria for obtaining

[[Page 28549]]

test subjects, and detail the protocols for administering the vision

tests.

6. Conduct empirical evaluation of operational feasibility. This is

the actual testing component, in which the drivers will be tested in an

operational setting to ensure that the new test battery's facilities

requirements are not excessive, its personnel needs are realistic, and

that the tests can be administered, scored, and interpreted in a timely

fashion by the individuals responsible for administering the tests.

7. Conduct empirical evaluation of validity of pass/fail criteria

for those candidate vision tests without sufficiently demonstrated

construct validity. The first step in this task would be to define the

study sample. The most likely sample would be age-matched 'visually

impaired' and 'visually unimpaired' subjects on the candidate tests

(all subjects would be required to hold a valid CDL). The researchers

would then conduct the vision and performance tests and analyze the

differences in performance between the two groups. Differences would be

measured relative to alternative cutoff scores, so that the Agency

could determine the significance of choosing different levels of

stringency.

The FHWA could decide to bypass task 7, the empirical validation

phase, if it determines that enough information currently exists to

establish a new standard, or that additional research would be unlikely

to lead to significant safety improvements. The Agency estimates that

skipping task 7 would reduce the time needed to complete the research

by one year.

8. Recommend tests and pass/fail criteria. Based on the work

completed above, the researchers would propose specific tests to be

added to the existing testing battery and cutoff scores for each test.

9. Convene interest groups to develop operational recommendations.

These groups would include motor vehicle licensing administrators,

researchers, industry associations, and safety advocates.

Proposed Long-Term Research

10. Design of tests and protocols. This includes developing the

visual performance and behavioral tests specified in task 2, generating

initial data from a pilot test, and designing draft protocols to be

used in later stages.

The visual function test would include some combination of existing

and new tests. The Agency would arrange the practical testing aspects,

including the purchasing and licensing of tests, acquiring any software

and documentation required, and developing the test protocols.

Pilot tests would be conducted on a small sample of drivers to

verify test reliability and suitability for large scale testing. The

Agency would modify procedures and protocols as appropriate. Upon

completion of the pilot test, the FHWA would conduct the visual

function tests on a medium-sized sample of drivers. The sample would be

large enough to allow the Agency to analyze test score characteristics.

Use of pilot tests would allow the Agency to ensure that a test would

produce useable results. In addition, correlation between tests may be

observed, in which case some tests may be eliminated from the final

battery as redundant.

For driver behavior measures, the agency would develop simulator

materials and closed-course testing procedures. To the extent possible,

the FHWA would employ procedures which can be used on multimedia

personal computers with a minimum of special equipment. The Agency

would develop hardware, software, and testing protocols.

The extent of the work performed in this task will depend on

whether the Agency conducts the short-term research. Some of the work

outlined above may be conducted in task 2 of the short-term track. If

that is the case, we will not repeat the work in this task.

11. Laboratory Simulation. This phase consists of evaluating the

candidate tests in a controlled setting, to identify and correct any

problems in the testing or protocols. This step is essentially a

``dress rehearsal'' for the full scale test. Because the next phase is

the most costly in terms of time and resources, this phase was designed

to allow the Agency to make a final decision about whether to continue

with the research prior to commencing with the next phase of the plan,

the full testing and evaluation.

A limited number of subjects would be given all the proposed visual

function and driver behavior tests. The results of these tests would be

analyzed extensively, including relationships between and among both

sets of tests. The analysis would address the following issues:

a. Are the distributions of scores useable?

b. Is there sufficient variance to discern relationships between

visual and driving tests?

c. How well do visual tests predict driving results, by themselves

and in combination with other tests? How much of the variance between

individuals in driving behavior can the vision tests explain?

The agency would also conduct a preliminary cost-benefit analysis.

In addition to projecting the cost of the next phases of the research

program, the analysis would estimate the cost of implementing a new

vision standard and the possible safety benefits.

12. Validation Testing. This phase consists of two sequential

activities, test preparation and data collection, the crux of the

proposed research.

Test preparation includes selection and configuration of test

sites, plus selection and preparation of subjects. The site (or sites)

selected must have, or be able to accommodate, a driving simulator, a

closed test course, and a road test course. Site preparation includes

configuring the testing equipment for the site, surveying the road test

course, and preparing and deploying signs and obstacles for the closed

test course. Preparation of the subjects consists of briefing the

participants and pre-testing them for the visual measures.

A final closed-course pilot test would then be conducted, using a

small number of drivers. This would provide the Agency with a final

opportunity to modify the test procedures.

Validation testing would probably include at least two distinct

activities, simulation and closed-course testing, and would possibly

also include controlled road testing. Variables would be strictly

controlled in these simulation tests to ensure the accuracy and

reliability of results. The FHWA expects that the simulators used for

this phase would be more sophisticated, with higher video resolution

than those used in the previous pilot test.

Closed-course testing would be used to test drivers under low

visibility conditions. This is difficult to imitate on a simulator and

is unsafe to test on the road.

If road testing is conducted, it would consist of non-intrusive

instruments to record driver responses, such as eye movement patterns,

blink rates, pupil diameter, and fixation points. This information,

combined with data on the roadway obstacles, provides a stream of data

related to working field of view, detection time, and how drivers react

to critical events. The road test would be conducted under normal

driving conditions to assure that the results are generalizable to

normal CMV operating practices. There are a number of hurdles to using

a road test, including the need to perform the test for an

impractically long period to obtain sufficient data, and the

possibility that drivers would modify their behavior if they are aware

[[Page 28550]]

that they are under observation. The FHWA would decide whether to

conduct the road test after analyzing the results obtained in the

simulation and closed-course tests.

13. Standard Development. The results of the preceding task would

be analyzed for validity, reliability, and practicality. If the results

of the validation testing justify specification of a new standard, a

decision framework for that standard would be constructed.

Specific Questions

The FHWA is specifically interested in comments addressing the

following issues:

1. Are there any methodological shortcomings in the research plan

outlined above that need to be addressed?

2. Is the plan likely to meet the objective of leading to an

improved, performance-based vision standard?

3. Does the plan reflect an understanding of the current literature

and consider its implications?

4. Is the plan capable of adequately addressing practical matters,

such as the cost of any new testing machinery developed, the level of

training required to conduct new tests, and the time needed to take

tests?

5. Has this type of research been conducted in other professions?

What were the results?

6. Should the FHWA proceed with the short-term plan, the long-term

plan, both, or neither?

7. Should the FHWA proceed with an alternative plan? If so,

describe that plan.

Current Status of the Research Program

The FHWA is currently in the midst of step 2 of the research plan,

which consists of inventorying existing tests and evaluating them

against a number of criteria, including their cost, which visual

functions they measure, overlap between different tests, and the amount

of training required to conduct the tests.

Format of Public Hearing

The FHWA announced in the November 17th notice (59 FR 59386) its

intention to hold a public hearing to discuss the research plan. The

public hearing will be held on August 9, 1996, at the Chicago O'Hare

Marriot, 8535 West Higgins Road, Chicago, IL 60631, (312) 693-4444. The

hearing will begin at 8:30 a.m. and conclude at 4:30 p.m.

Individuals wishing to speak at the hearing should contact the FHWA

at the address or phone number listed above under the heading ``For

Further Information Contact.'' Individuals may submit written comments

in addition to, or in place of, oral testimony. All commentors will be

limited to ten minutes of oral remarks.

The hearing will commence with an explanation of the proposed

research plan, including a brief description of the background to this

effort, the goals of the proposed research, and the steps of the

proposed plan. The FHWA will then accept questions from audience

members, with individuals who have contacted the FHWA given the first

opportunity to speak.

(49 U.S.C. 31136(a)(3), 31502)

Issued on: May 20, 1996.

Rodney E. Slater,

Federal Highway Administrator.

[FR Doc. 96-14041 Filed 6-4-96; 8:45 am]

BILLING CODE 4910-22-P