Category: Blog

Tapping the Well – Using PubMed

There is much medical information can do to enhance your practice of law. Whether you are a personal injury attorney, a criminal defense attorney, or practice in the area of products liability you will often have a need to research issues that are discussed in the medical literature.

You do not have to be a professional librarian to find the information you are seeking. This article will give you a brief outline of how to do a simple search on PubMed as well as explain some of this database’s special features.

PubMed was developed by the National Center for Biotechnology Information at the National Library of Medicine, located at the National Institutes of Health. It was developed in conjunction with publishers of biomedical literature as a search tool for accessing literature citations and linking to full-text journals at Web sites of participating publishers.

Publishers participating in PubMed electronically supply the National Library of Medicine with their citations prior to or at the time of publication. If the publisher has a Web site that offers full text of its journals, PubMed provides links to that site. PubMed is updated weekly.

PubMed is free to anyone with Internet access and provides access to two main databases, MEDLINE and PreMEDLINE. MEDLINE is the National Library of Medicine’s bibliographic database covering the fields of medicine, nursing, dentistry, veterinary medicine, the health caresystem, and the preclinical sciences. The file contains approximately nine million records dating back to 1966, and covers bibliographic citations and author abstracts from approximately 3,900 current biomedical journals. PreMEDLINE provides citation and abstracts before the citation is indexed with MeSH headings. These records are added daily to PreMEDLINE and appear in PubMed with the tag: [Medline record in progress].

Basic Search Strategy

Pick a search topic, for example: Is chickenpox vaccination harmful to children? First divide the topic into concepts, then find keywords or subject headings (MeSH) to represent these concepts. Finally combine concepts with the Boolean Operators AND, OR, or NOT. The Boolean Operators must be capitalized in a search. These operators are used when searching for more than one concept or search statement

Example: “vaccination OR vaccine” searches for records containing either term (OR) and will retrieve a greater number of records than using each term separately.

Example: “chickenpox AND vaccination” searches for records containing both terms (AND) and will retrieve a fewer number of records than using each term separately

Example: “vaccination NOT rubella” eliminates terms in one search set from a previous set (NOT) and will retrieve fewer records than using the first term separately.

If you use two or more Boolean Operators (AND, OR, NOT) in one search statement, the one on the left is processed first.

Example: “chickenpox AND vaccine OR vaccination” would retrieve articles that have both chickenpox and vaccine and it would also retrieve all articles on vaccination (pneumonia, measles, etc)

To change the meaning use parentheses.

Example: “chickenpox AND (vaccine OR vaccination)” this would retrieve articles that have both chickenpox and either vaccine or vaccination, which would be fewer records than the example above.

You can use Truncation to find the first 150 variations of the root word. Truncation uses the symbol * as a substitute for zero or more characters.

Example: risk* retrieves risk, risks, risk-taking. However the term infection* would not get infection control because there is a space after the word infection.

Another way to search is by using logical terms. To be sure the phrase is recognized use ” “

Example: “Vaccination complications”

Simple Search Strategy Techniques
Enter your search term(s) in the “Search Box” (see below) The terms are matched against a MeSH translation table, Journals translation table, a phrase list and an author index. Use the “Enter” key or click on “Go” to get a display of your results as brief citations in batches of 20 by default. To change the number of items displayed, select a higher number from the pull down menu “Show”. To change the Display format, select from the pull-down menu next to Display.

To search by an author’s name enter the name in the format of last name plus initials (no punctuation)

Example: smith ja

PubMed automatically truncates on the author’s name to account for varying initials and designations such as Jr. or 2nd. This prevents retrieval by smith a, smith b, etc. A name entered with this format will search in the author field. If only the author’s last name is entered, PubMed will search that name in All Fields, not just the Author field. An exception to this is when the author name is found in the MeSH (National Library of Medicine’s Medical Subject Headings) Translation table (e.g., Yang will search as Yin-Yang [MeSH] or Yang [Text Word].)

Note: Use double quotes around the surname and first initial qualified with the author search field tag [au] to turn off the automatic truncation and retrieve based on only the single first initial.

Example: “smith j” [au]

Journal Titles may be entered in full:

As valid MEDLINE abbreviations:

Or as ISSN numbers (a standard number for journals), e.g., 1059-1524:

After you enter your search, your results will be displayed. The display format may be changed by selecting form the pull down menu by the “display” button.

Your Citations can be displayed in any of the following formats.

Summary – This format may include: Author name(s), Title, Journal Source, Record Status (for in-process or publisher-supplied citations), indication if article is non-English, Publication Type for review or retracted publication, “No abstract available” notation, and PubMed and MEDLINE Unique Identifiers.

Brief – Author name, first 10 characters of the title, and the PMID.

Abstract – This format may include: Journal Source, record status for in-process or publisher-supplied citations, Title, indication if article is non-English, Authors, Author Affiliation, Abstract (if present), Publication Types (except for the Journal Article publication type), Personal Name as Subject, Erratum, Retraction,Comments, and PubMed and MEDLINE Unique Identifiers.

Citation – This format may include: Journal Source, record status for in-process or publisher supplied citations, Title, indication if article is non-English, Authors, Author Affiliation, Abstract (if present), Publication Types (except for the Journal Article publication type), Erratum, Retraction, Comments, MeSH Terms, Personal Name as Subject, Chemical Substances, SI-databank accession numbers, Grant numbers, and PubMed and MEDLINE Unique Identifiers.

Advanced Searching Techniques

Using the “Limits” feature:

Click Limits from the Features bar to limit your search to a specific age group, gender, or human or animal studies. Limits also allows you to restrict to articles published in a specific language, and to specific types of articles such as review articles. You can limit by either Entrez or Publication Date. And lastly, you may limit your retrieval to a specific subset of citations within PubMed, such as AIDS-related citations or in-process citations, i.e., PreMEDLINE.

If you select a limit and either run a search or move to another screen, a check will appear in a box next to Limits on the Features bar to indicate that limits have been selected. If you then run a search, the limits in effect will appear in the yellow bar above the Display button. To turn off the limits before you run your next search, click on the box to remove the check (see below)

You may use this feature to limit your retrieval to only citations that contain an abstract by using the check box next to only items with abstracts. Citations to articles published prior to 1975 do not include abstracts.

All citations in PubMed are for journal articles. However, you may limit your retrieval based on the type of material the article represents (e.g., Clinical Trials or Review articles). The Publication Types pull-down menu contains a list of frequently searched publication types. If you do not make a selection, PubMed will not restrict to any particular publication type.

PubMed indexes journals published in approximately forty languages. The Languages pull-down menu contains a list of frequently searched languages. If you do not select a language, PubMed will not limit your retrieval by language.

PubMed contains citations published back to 1966 and new citations are added daily. The Entrez Date is the date the citation was added to PubMed. The Publication Date is the date the article was published. When PubMed displays your search results, the citations are displayed in descending Entrez Date order i.e., last in, first out. If you do not specify a date range, PubMed searches for citations to articles back to 1966.

Use the Entrez Date pull-down menu to limit your search back in time from 30 days to 10 years. The Publication Date pull-down menu toggles between Publication Date and Entrez Date.

Using the Citation Matcher

The Citation Matcher options allow you to find the citation or the PubMed ID of any article in the PubMed database using bibliographic information Use the Single Citation Matcher to look for a single citation. This feature is a fill-in-the-blank form that lets you enter journal citation information to locate a single citation, or items from a particular volume or issue of a journal.

The Batch Citation Matcher allows you to retrieve the PubMed IDs for many articles all at once. This feature requires that you enter the bibliographic information (journal, volume, page, etc.) in a specific format.

If you wish to match citations in bulk by e-mail, send email to [email protected] For instructions, send an message which is empty except for the word “help” in the subject or body.

A list of all journals that are included in the PubMed database is available by FTP in the GNU Zip, Uncompressed, UNIX Compress, or PKZIP format.

Using the “Preview/Index” Feature

The Preview/Index feature is used to preview the number of search results before displaying the citations, refine search strategies by adding one or more terms one at a time, add terms to a strategy from specific search fields, view and select terms from the Index to develop search strategies.

To preview the number of results before displaying the citations, type your term(s) in the query box and click Preview. Preview displays the number of citations in your search results. To refine your search strategy, add another term to the existing term(s) in the query box and click Preview. The additional terms will be combined with the existing terms, and the new search with the new number of citations will display. Continue adding terms until your strategy is complete. To display your results, click on the result number (hypertext) in the Preview display.

You can also use Index to select from a list of terms within a search field. Select a search field from the All Fields pull-down menu, enter a term in the box, and click Index. PubMed displays an alphabetic list of terms in the Index for the selected search field. The number of citations in PubMed that contain the term appears in parentheses to the right of the term. You can scroll up or down the list, or click Up or Down to move along in the Index.

Highlight a term by clicking on it. Then click the appropriate operator, AND, OR, NOT. The selected term will be added to the query box. To OR together multiple terms from an Index display and then add (or AND) them to your search, click on each term while holding down the Ctrl-key(PC) or the Command-key(Mac). When all the terms you want are highlighted, click the connector AND to add the terms to the query.

You may then continue to add additional terms from other search fields. Once your strategy is complete, click Preview to see the number of results, or click Go to display the citations.

For example, to see the MeSH Terms beginning with “chickenpox,” select MeSH Terms from the All Fields pull-down menu, enter the term, chickenpox, in the text box, and click Index. Scroll down in the Index box, highlight the MeSH Term “chickenpox vaccine,” and click AND. “Chickenpox vaccine” [MeSH Terms] will be added to the query box.

Using the Related PubMed Articles Link Each citation in PubMed has a link that will retrieve a pre-calculated set of PubMed citations that are closely related to the selected article. Click on Related Articles to the right of each citation to display the related set of articles. PubMed creates this set by comparing words from the title, abstract, and MeSH terms using a powerful word-weighted algorithm. Citations are displayed in rank order from most to least relevant with the linked from citation displayed first.

Using the History Feature

PubMed will hold all your search strategies and results in the History. You can see your search History by clicking on History from the Features bar. History is only available after you run your first search. History lists and numbers your searches in the order in which you run them. The History displays the search number, your search query, the time of search, and the number of citations in your results. To view the results from a search, click on the number of results.

You can combine searches (see below) or add additional terms to an existing search by using the pound sign (#) before the search number, e.g., #2 AND #6, or #3 AND (drug therapy OR diet therapy). Once you have entered a revised search strategy in the query box, click Go to view the search results. Click Clear History to remove all searches from the History screen and the Preview/Index screen.

The maximum number of queries held in History is 100. Once the maximum number is reached, PubMed will remove the oldest search from the History to add the most current search. .The Search History will be lost after one hour of inactivity on PubMed or any of the other Entrez databases. The Clear History button in History will also clear the Preview/Index history information.

PubMed will move a search statement number to the top of the History if the new search is the same as a previous search. A separate Search History will be kept for each of the Entrez databases although the search statement numbers will be assigned sequentially for all databases .PubMed uses cookies to keep a history of your searches. In order for you to use this feature your web browser must be set to accept cookies .Citations in the Clipboard are represented by the search number #0 which may be used in Boolean search statements. For example, to limit the citations you have collected in the Clipboard to English language citations use the following search: #0 AND english [la]. This does not affect or replace the Clipboard contents.

Saving your search results

To save the entire set of search results use the pull-down menu to select the desired format and then click Save. To save specific citations, click on the check box to the left of each citation (you may move to other pages in the retrieval), when you have finished making your selections, click Save.

The Save option (from the search results page) will save the entire set of search results even if only the first batch of citations are displayed. The maximum number of items that can be saved is 5000. If you try to save a file with more than 5000 citations, PubMed will display a message that instructs you to refine your search. The default file name is query.fcgi. Consider changing the name to something more meaningful to you, and the .fcgi extension to .txt, if you wish to open the file in a text editor or word processing package.

To save citations in HTML format use the Save as… function of your browser. Change the file extension to html. You will only save the citations displayed on the screen so you may wish to use the Show function and Text button to adjust your display as needed.

Printing your search results

Use the print function of your Web browser to print all the information and citations displayed on your web page. Before printing, consider using Show to increase the number of documents per page so that the total number of documents is displayed on one page (maximum: 500 per page). You can only print the citations from the displayed page. You may also wish to display your citations as Text to strip the sidebar menu and toolbars prior to printing your results.


The Clipboard gives you a place to collect selected citations from one search or several searches. You may then want to use the print, save, or order buttons. The maximum number of items that can be placed in the Clipboard is 500. Once you have added items to the Clipboard, you can click on Clipboard from the Features bar to view your selections. Citations in Clipboard are displayed in the order they were added.

Saving from the Clipboard

Citations are initially displayed in the summary format in descending Entrez date order. You can select all or individual citations to display or save in one of the six display formats. Select the desired format from the pull-down menu, click Save to save your selections to a file, or use the Print feature of your web browser to print the citations. Printing from your web browser will only print the information and citations listed on the web page. You may also display citations as plain text without the sidebar menu and toolbars by clicking on the Text button.

Your results may be on more than once page. PubMed will retain your selections from all pages of your results. To mark specific citations to save, click on the check box to the left of each citation and continue to page through your results. Use Save to save all the selected citations to a file.

Ordering full text copies of journal articles

Order allows you to order the full-text copy of an article from a library in your area using the Order Documents (Loansome Doc) feature of PubMed. Prior to using this program you must establish an agreement with a Loansome Doc participating library. Local fees may apply. For information on the medical libraries in your area (or country) that can provide the Loansome Doc service, click on Order Documents from the PubMed sidebar menu, then “Registration” from the Loansome Doc Ordering System screen. Once you have registered with Loansome Doc, you can order citations, by clicking on the check box to the left of each citation. When you have finished marking your selections (you may move to other pages within your results), click Order. You may also use the Clipboard to collect items from multiple searches before ordering.

For some journals the full-text of articles are available via a PubMed link to the publisher’s Web site. Publisher links for the full-text of the article are displayed on the Abstract or Citation display. You may also click on LinkOut to the right of each citation in the abstract, citation or MEDLINE display. LinkOut is a PubMed service that links to outside sources for the full-text of the article, e.g., a publisher’s web site, as well as other resources such as biological databases, and sequencing centers. User registration, a subscription fee, or some other type of fee may be required to access the full text of articles in some journals.

University of Oklahoma Law Center Commencement Address – Division of Legal Assistant Education Graduation

“We cannot direct the wind, but we can adjust the sails.”

The beauty of the law is that it is dynamic – it changes as society changes. The law adapts to technological developments, to shifts in social values, and to the problems of modern society. The key to succeeding in this dynamic legal environment is that we too must learn to adapt-to adjust our sails.

As I look at all of you today, I cannot help but see myself ten years ago, a new graduate of this highly respected program. Back in 1993, I was working part-time in a law office while teaching nursing full time, with a few weekends of hospital work thrown in the mix to keep in touch with the “real world.” The rest of my “free time” was spent with studying and attending classes here. In anticipation of graduation, and a move out of state to Texas…I sent resumes to quite a few law firms looking for the perfect place to begin my new career. I interviewed with several. The first firm I spoke with kept their paralegals in a quansit hut near the back parking lot, isolating them from the main activity of the firm, I did not feel that would be a good place to work, and learn…so I continued interviewing and searching. One firm stood out in my job search. This firm was over 100 years old, and had attorneys with very different backgrounds and experiences. The combined experience of these five attorneys spanned nearly 75 years practicing law. Often the partners and their secretaries would share stores of their challenges, clients and, and even failures. Although I did not realize it at the time, working for this firm would be a life-changing and career developing experience for me.

I worked with five different attorneys, each with his own style and personality. Very early I learned that part of my job was to adapt to their differing approaches and preferences. I was often “adjusting my sails” to accommodate them. I was asked to work in areas of law that were unfamiliar to me. I had to learn to listen, to ask questions, to think creatively, to accept criticism, and to adjust my sails.

Some of the most important lessons I have learned over the years were not ones from a textbook. And today I will share them with you. Among these “real world” lessons were:

Most law offices are not like the ones depicted on shows like The Practice or Ally McBeal, and most attorneys are not going to look or act like Dylan McDermott or Calista Flockhart.

Do not assume, always verify the facts.

Pay attention to detail. No one ever lost a case by knowing the facts too well.

If you don’t know the answer, know the resources you can use to find it. Knowing the resource is half the battle.

Proofread all work. One mistake can compromise your case. Even though I didn’t like it, I learned this lesson-sometimes the hard way-but I learned it.

Respect your attorney and their client by meeting or exceeding deadlines. No one has ever been fired for finishing a project early.

Networking is invaluable. You can never know too many knowledgeable people.

Treat legal secretaries with respect. You will find that they actually know more about some things than you do.

Listen and learn. Everyone has something to offer. You can learn from the most experienced of veterans and the most inexperienced “greenhorns.” Great ideas sometimes come from the most unexpected sources.

Document, document, document…each task completed, each client communication, each conference with an attorney and all deadlines.

Do not participate in office gossip, in the end it serves no purpose and is a distraction from your assigned tasks.

Do not lose sight of who you are really working for: the client. You may have twenty files On your desk, but to each client, their file is the only one that is important to them.

Join and participate in local, state and national professional organizations, and take advantage of continuing education programs offered to their members. Through these activities you can exchange ideas, learn new resources, and develop a valuable network of peers.

I believe all of these “real world” lessons are important, but the most important lesson I have learned is that success can only be achieved on your terms. Everyone has his or her own idea of what a successful career is. Ultimately, true success is how each of us chooses to define it. When I started this program, my friends and family asked me why? I had an education, and a career, was I crazy? When I left a 10-year career in nursing education to work for attorneys, my nursing colleagues thought I was crazy. When I changed careers again to be an independent medical legal consultant, they again said I was crazy. But I am not crazy. I am happy. I am successful on my own terms. Don’t let someone else decide what your success should be. If your success is not on your own terms-if it looks good to the world but does not feel good in your heart-it is not success at all.

The wind changed many times in my life and each time I adjusted my sails. My career has been an incredible journey. Never in my wildest dreams would I have imagined in 1993 that I would be standing here ten years later, before a graduation class of my peers, and telling of my experiences and life lessons, much less giving career advice.

Each class, each job, each life experience along the journey was a foundation for the next step. Today you take a step toward your careers as a paralegal. Learn from each case you assist with. Learn from each attorney you work with. Let each of these lessons be a foundation for your next experience. And, most importantly, when the wind changes adjust your sails. An incredible journey awaits you.

Why Injuries Occur in Minor Vehicle Damage Collisions

The reasons why injuries can occur in collisions with little or no damage have been examined. Analysis of the underlying math and physics reveals that reliance on vehicle damage alone to determine the severity of a collision is invalid. Discounting injuries based on lack of vehicle damage is not supported by either the laws of physics or the empirical data available. Additionally, the use of a threshold for injuries has been refuted by the principles of biomechanics and the empirical data.

The reconstruction of many traffic collisions is significantly complicated by the absence of hard data. It is not uncommon for the accident either to be not investigated, or investigated only superficially. This is especially true in collisions where the property damage is minor. Compounding the problem is that many of these collisions appear to be noninjury crashes until a few hours after the police have left 1. Often, the only information available is the damage to the vehicles, the statements of the parties involved and the medical data. A thorough analysis of both the speeds involved and the injury potential of the collision requires the inclusion of all three of these items. However, there is not a quantitative link between the speeds involved and the injuries received. As such, speed determination and injury potential can be discussed separately.

PART1: Speed Determination in Minor Vehicle Damage Collisions

Speed determination of vehicles can be both qualitative and quantitative. Qualitative is usually subjective, and involves personal observations of witnesses or participants. Examples of this include statements about relative speed, “I was traveling at 20 m.p.h. and the other car was moving at the same speed”, or statements about absolute speed; “The car was going about 55 m.p.h.”

Quantitative determination of the speeds in traffic collisions involves the determination of a precise number or range of numbers and can be accomplished in two ways. The first is direct measurement. This includes the use of items such as RADAR guns and lasers. However, it is rare that direct measurement of speed exists in actual collisions.

The second quantitative method in speed determination is the application of the laws of physics to the physical evidence at the scene. This can include damage to the vehicles, preimpact evidence and post impact movement of the vehicles.

The quantitative method has significant limitations in minor vehicle damage collisions. Often there is minimal physical data on preimpact action or post-impact movements. For this reason, it is often necessary to combine quantitative physical evidence and the qualitative observations of the participants regarding actions and movement.

The most common qualitative attempt made is to look at the damage to the vehicle and attempt to divine an impact speed. From this, changes in velocity (delta-V) can be inferred 2. Determining an impact speed from damage in collisions where there is little or no damage has severe limitations. If a collision results in no visible damage, and the impact is bumper-tobumper, it is probable that the impact speed was over 10 m.p.h. This is based on hundreds of crash tests that have been run throughout the United States, Canada, Europe and Japan i, ii, iii, iv, v, vi, vii, viii. As an example, Figure 1 shows the rear bumper of a Ford Festiva that was just hit from behind at approximately 11 m.p.h. This vehicle had no damage despite the fact it had been involved in a frontal impact at approximately 7 m.p.h. previously that day.

The qualitative method is also limited when the reconstructionist has only pictures of a vehicle, often taken in such a way that damage is not visible. Figure 2 shows a Dodge Intrepid that was just impacted at over 20 m.p.h.

Even with repair estimates it is difficult to ascertain anything less than a minimum impact speed. If damage is visible, a reasonable minimum is a 10 m.p.h. impact speed. If damage extends beyond the bumper, 15 m.p.h. is a reasonable minimum impact speed. It is rare to have an impact speed as high as 20 m.p.h. with no damage. However, the absence of visible damage in a poor photograph is not the same as there being no damage.

Another source of information that can be used, and is often overlooked, is frame or unibody damage. This is often the first damage that is visible in a rear impact and is captured in the repair estimates only a fraction of the time. It can be inferred through physical inspection of the vehicle and proven with inexpensive laser measurements. It is usually not visible in photographs taken by insurance companies. Figure 3 shows a vehicle with significant unibody distortion that required physical inspection to identify it. Figure 4 is a printout showing the unibody damage. This vehicle had distortion of 17 millimeters that was not visible in the photographs.

The common error in determining speed based on damage where there is little or no damage is to make an invalid comparison to a non vehicle-to-vehicle crash tests. The most widespread of these is comparison to tests run by the Insurance Institute for Highway Safety (IIHS). Almost as widespread, is the comparison to bumper standards. Finally, comparison to other non-IIHS crash tests is common.

IIHS Crash Tests
Every year the IIHS, an insurance industry funded group, runs numerous crash tests of vehicles into barriers at low speed. The IIHS then publishes the results as an indication of how much damage a vehicle will incur in a low speed collision. Usually these tests are run at 5 m.p.h. into walls or poles. Many times a reconstructionist will attempt to compare the damage in a vehicle-to-vehicle collision with the damage in an IIHS test. Based upon the principle of conservation of energy, this will grossly underestimate the actual collision speed (see the discussion of barrier impacts below). As an example, the Dodge Intrepid, (See figure 2) has a bumper rating of “marginal” from the IIHS, just above the lowest rating of poor.

Most midsize small Honda vehicles receive a bumper rating of “acceptable”. Figure 5 shows the Honda Prelude that struck the rear of the Intrepid in Figure 2. The impact was initially bumper-to-bumper. While the IIHS did not provide a specific rating on the Honda Prelude bumper, based on barrier test, the damage should be reversed.

IIHS barrier test data does not reflect the performance of actual bumpers in vehicle-tovehicle collisions. The data only provides information on how bumpers will perform when running into walls.

It is worth noting that the IIHS implies that the reason they run barrier tests is to ascertain the performance of bumpers in vehicle-to-vehicle collision. Their web site ix has the following statement;

Bumpers should protect car bodies from damage in low-speed collisions, the kind that frequently occur in congested urban traffic. But many don’t. The Institute assesses performance using the costs of repairing vehicle damage in a series of four crash tests at 5 mph — front- and rear-into-flat-barrier, front-into-anglebarrier, and rear-into-pole.

IIHS does not test bumpers under conditions that match their implied goal. Their data does not establish the actual performance of bumpers in vehicle-to-vehicle collisions. The data collected by the IIHS has no relevance in determining the impact speed in a vehicle-to-vehicle collision, although it is commonly used in litigation in support of the insurance companies that fund the IIHS 3.

Bumper Standards
A misconception often propagated is that if a vehicle has a 2.5 m.p.h. bumper and the bumper is not damaged, the impact speed was under 2.5 m.p.h. This is a misapplication of the bumper standards x. The standard establishes the conditions under which bumpers cannot be damaged. It does not establish the standard under which a bumper must be damaged. Empirical evidence reveals that in vehicle to vehicle collisions xi, the bumpers will not be damaged until bumper standard speed has been exceeded by a factor of 2 to 8 times.

It is worth noting that the design of modern bumpers often prevents the direct observation of bumper damage without physical removal of the bumper.

Non vehicle-to-vehicle impacts
The most common of these are barrier impacts of the type described in the section on IIHS. A vehicle 4 is run into a barrier of some kind and the damage is compared to the actual collision. However, the physics of the collision does not support this. In a barrier impact, virtually all of the energy goes into damaging the vehicle since it cannot go anywhere else. In vehicle-to-vehicle collisions, most of the energy goes into moving the vehicle. Therefore, a car that is damaged in a 5 m.p.h. barrier test, may not be damaged in a 15 m.p.h. vehicle to vehicle collision xii.

Table 1 Damage comparison between VTB and VTV collisions.

Table 1 is a comparison of four Honda Accords from tests run by the National Highway Traffic Safety Administration (NHTSA). The first vehicle impacted a barrier. The other three vehicles were involved in vehicle-to-vehicle (VTV) collisions at significantly higher speeds. Even when the car-to-car collision had three times the energy of the vehicle to barrier collision (VTB), the damage was less. The percentage listed is relative to a barrier impact.

Computer Programs
Solving for the speed of a vehicle in an automobile accident often involves solving up to six simultaneous equations with over a dozen variables. For this reason, even when fewer equations and variables are involved, computers have been employed to help solve the equations.

Among the earliest attempts to solve this problem was the Simulation Model of Automobile Collisions, (SMAC) developed by Calspan under DOT contract. In this program, the initial conditions, such as speed, type of vehicle and directions were entered into the computer and the system would determine where the vehicles should end up. The investigator could then look at the results and alter the initial conditions until the solution matched the final results of the actual collision being investigated. The problem with this approach was the large amount of computer time required to run a single simulation 5. For this reason, a preprocessor was written in an attempt to get a starting point that was “close”.

The preprocessor was the Calspan Reconstruction of Accident Speeds on the Highway (CRASH) xiii. This program used several approaches in an attempt to approximate a collision. One method used the post impact movement of the vehicles, another used the damage to the vehicles while a third used both. This program was thought to be accurate to within 10% in collisions with a change in velocity of under 15 m.p.h. However, full scale testing revealed that in this range, the program isn’t reliably accurate to within 100% xiv and can actually return impossible answers.

Despite these shortcoming in the program, and the fact it was designed to be a first step in a reconstruction, the equations developed were copied and propagated throughout the accident reconstructioncommunity. Over time, the equations took many forms and became increasingly accepted as scientifically valid. Even when full-scale tests showed the equations to be invalid they weren’t abandoned, they were modified and adjusted. This occurred in spite of the fact that the basic underlying assumptions were not valid for anything more than a first order approximation 6.

In an attempt to validate the approach, there was a full-scale study done in 1978 called RICSAC xv. The study consisted of 12 staged collisions of various speeds and crash geometry. These crashes were used to determine the validity of the CRASH program and its successors. An attempt was made to define the error of the estimate independent of impact speed. The problem with this approach is that the impact speed is in error if it deviates from the measured data from the RICSAC tests. Table 2 is a graphical representation of the RICSAC results 7. The table shows that for all cases, the calculated values are different from the measured values. In one case, the program overstated the actual velocity of the vehicle (measured at 31.5 MPH) by over 20 MPH and in another instance shows the vehicle going backwards at 6 MPH when in reality the vehicle was not moving. Another case showed the calculated vehicle speed (measured at 31.5 MPH again) was understated by 11 MPH.

Numerous other tests have been run in an attempt to validate the model and all have shown unacceptable errors. In fact, the User’s manual from CRASH3 actually warns that the results may have no significance for any single collision. The problem this generates is that there is no mechanism for knowing when the program has given a correct answer.

Numerous other programs exist, some based on the CRASH3 model and some based on attempting to determine the energy involved in damaging the vehicle. The latter usually involves the determination of very low energy values for the damage. The problem with this is that thousands of ft-lbs of energy can be absorbed by the vehicles with no visible damage. This is complicated by the fact that energy absorption in collisions with no visible damage does not follow an identified pattern.

Other Errors
Other errors that are often seen but will not be extensively discussed include the use of bumper isolators and photographs to determine speeds and ignoring foam core bumper damage. Isolators are often used to assert a maximum impact speed. However, they only have some applicability in determining minimum impact speeds.

Photographs will also only provide reasonable minimum speeds. Figure 6 shows a typical low quality photograph often used to determine impact speed. The difficulty with this approach is that it is often hard to actually observe the damage.

This is informative because frequently the reconstructionist is only provided poor photographs from the insurance company. In “A Guide for Risk Managers and Claims Personnel 8”, the following statement appears.
“The most common mistake for an accident investigator is to just take a couple of overall shots of the vehicle, at some unknown and oblique angle.”
However, this is the most common type of photograph provided by insurance companies.

Foam core bumpers:

Foam core bumpers often show no visible damage even when the steel beam in the center is damage. In order to see this damage, it is common that the bumper must be removed. Figures 8 shows a foam core bumper with no visible damage. Figure 9 shows the damaged bumper of the vehicle in figure 8.

Part 2: Injury Potential in Minor Vehicle Damage Collisions

  • Location of applied forces
  • Magnitude of applied forces
  • Daily Activities
  • Injury Databases versus Injury Thresholds
  • Injury Study Data
  • Aggravating Factors

John Smith is the president of Raymond P. Smith and Associates. He has published and lectured extensively in the area of accident reconstruction and biomechanics. Mr. Smith was the supervising engineer for the low speed crash test depicted in the video “Four Speeds”.

The delay of symptom onset has been well documented in other articles. The Association for the Advancement of Automotive Medicine has reported that over 50% of injuries are not identified for 6 hours. Full scale testing has also revealed this delay.

Speeds can be calculated using the law of conservation of momentum and the coefficient of restitution.

Due to the principle of conservation of energy, strengthening bumpers actually results in increased injury potential in lower speed collisions, opposite of IIHS’s implied goals.

Occasionally vehicles are modified from manufacture’s specifications before the tests are run, altering the damage pattern.

The program was written before the days of desktop computers when everything ran on a main frame.

The derivation of the equations in the program required the use of numerous simplifying assumptions. However, when running the program, these assumptions are violated, invalidating the equations and any results.

Analysis showed that R squared was .5752. Statistically this means the calculated results from the program are meaningless. CRASH3 was a later version of CRASH.

“A Guide for Risk Managers and Claims Personnel” A Publication of Ruhl & Associates

“Investigation of the Kinematics of Whiplash”, Mertz, Patrick, SAE Paper 670919, 1967

Human Occupant Kinematic Response to Low-Speed Rear End Impacts, Szabo, Welcher, SAE 1994

An Investigation into Vehicle and Occupant Response Subjected to Low Speed Rear Impacts, Navin, Romilly, Proceedings of the Multidisciplinary Road Safety Conference VI, June 1989

Human Subject Kinematics and Electromyographic Activity During Low Speed Rear Impacts, Szabo, Welcher, SAE 1996

Engineering Report on Impact Tests, Smith, July 1997

Characteristics of Specific Automobile Bumpers in Low Velocity Impacts, SAE 940916

Analysis of Human Test Subjects Kinematic Responses to Low Velocity Rear End Impacts, McConnell, Howard, Guzman, Bomar, Raddin, Benedict, Smith, Hatsell, SAE 1993

Head/Neck Kinematic Response of Human Subjects in Low-Speed Rear-End Collisions, Siegmund, et al, SAE 1997

Federal Motor Vehicle Safety Standard 215


“Damage Only, It Doesn’t Work; Minor Vehicle Damage Doesn’t Mean No Injury”, John J. Smith, Alan J. Mencin

“User’s Manual for the CRASH Computer Program”, McHenry, R.R; February 1976, NTIS #PB 252115

“Further Validation of EDCRASH Using the RICSAC Staged Collisions”, Day, T.D., and Hargens, R.L., SAE Paper No. 890740.