A Multidimensional Approach to Pain Assessment Using The BHI™ 2

©2004 by John Mark Disorbio EdD and Daniel Bruns, PsyD

BHI™ 2 ©2003 by Pearson Assessments

All Rights Reserved.

Last update 4/12/2004

Pain is one of the most common reasons why persons schedule a visit with their physicians. Indeed, research has shown that the cost of the treatment of chronic pain exceeds the costs of treatment of other disorders, including cancer and heart disease (Fishman, et al, 1997).  However, while cancer and heart disease are conditions for which there are objective measures, pain is a much more intangible condition. 

To understand this problem, one must begin with a definition of pain.  The International Association for the Study of Pain (IASP) defines pain in the following manner:

Pain: An unpleasant sensory and emotional experience associated with actual or potential tissue damage, or described in terms of such damage.

Note:  Pain is always subjective.  Each individual learns the application of the word through experiences related to injury in early life.  Biologists recognize that those stimuli that cause pain are liable to damage tissue.  Accordingly, pain is that experience we associate with actual or potential tissue damage.  It is unquestionably a sensation in a part or parts of the body, but it is also always unpleasant and therefore also an emotional experience... Many people report pain in the absence of tissue damage or any likely pathophysiological cause; usually this happens for psychological reasons.  There is usually no way to distinguish their experience from that due to tissue damage if we take the subjective report.  If they regard their experience as pain and if they report it in the same ways as pain caused by tissue damage, it should be accepted as pain... Activity induced in the nociceptor and nociceptive pathways by a noxious stimulus is not pain, which is always  a psychological state, even though we may well appreciate that pain most often has a proximate physical cause. (Mersky and Bogduk, 1994)

This definition leads to several remarkable insights.  First of all, there is the idea that Pain is not just a sensation, but is also an emotion.  At first this may seem surprising, but on second thought, it makes a good deal of sense.

Consider the phrase in the definition of pain that “Each individual learns the application of the word [pain] through experiences related to injury in early life.” Think of a toddler who is learning to run for the first time, and who slips and falls on a rock. The child skins an elbow, and runs to the parent crying in pain. What is being felt? It is an indivisible conglomerate of nociceptive pain sensations, anger at the stupid rock, and fear arising from seeing one’s own blood and the resultant feelings of vulnerability. The same principles can be extended into adulthood, and can be readily seen in a worker with a painful injury who is angry with the employer, or the patient who is terrifed by cancer-related pain. The sensation of pain and the associated emotional distress thus combine to form suffering, an undifferentiated mass of bad feelings. Patients, however, often fail to distinguish the various components of their suffering experience, and may simply label it “pain” (Fordyce, 1988).

The IASP defintion also states that pain is not to be equated with "activity induced in the nociceptor and nociceptive pathways by a noxious stimulus." "Nociceptors" are sensory receptors that detect real or potential tisue damage, while the "nociceptive pathway" is the path through the nervous systme that this information follows. The activation of such nerve by a stimulus cannot be equated with pain however, as pain is an experience. Research has shown that sometimes when these nerves are activated, the subject reports no pain. In other circumstances, subjects do experience pain even when these nerves are not activated.

What experiences should then be called pain? Using the IASP definition, pain is pain if the person looks at that way, whether or not there is any identifiable physical cause. Pain is thus a complex, sensory/emotional/cognitive experience that is subjective in nature.

The Assessment of Pain

In physical rehabilitation, there are many variables such as strength or range of motion that can be measured objectively.  Since pain is a subjective psychological state though, the only way of assessing it is through the patient’s verbal report.  However, verbal reports are known to be shaped by a variety of psychosocial variables.  Consequently, in order to assess pain reports thoroughly, one must know more about the person who is doing the reporting, and obtain a better understanding of the psychosocial context in which these reports occur.

Some may feel that since pain is a subjective phenomenon, it cannot be measured scientifically.  However, this is not true either. Psychometrics is the scientific method used for creating psychological tests, and is also the method used to assess subjective psychological states.  Although psychometrics is a field that most physicians are unfamiliar with, nevertheless, the assessment of pain depends on it.

In an extensive review of the literature (Meyer et al., 2001), it was determined that psychological tests are comparable in validity to medical tests.  Thus, even though subjective psychological states have an intangible quality, nevertheless, scientific measurement of them seems quite possible. Unfortunately, many of the most commonly used methods of assessing pain have significant psychometric difficulties. A good example of this is the Visual Analog Scale for measuring pain. 

The origins of the Visual Analog Scale (VAS) are unclear, but it came under scientific scrutiny about 20 years ago (Price, et al, 1983). Following this, variations of the VAS rapidly proliferated.  Within two years, one study compared 20 alternative versions of the VAS (Seymour et al., 1985). Since that time, varieties of the VAS have become too numerous to count, and there is no “official” version of it.

Although the VAS is a standard means of assessing pain in medical settings, it is not standardized.  Using an unstandardized measure of pain is like using a thermometer, without knowing if it is Fahrenheit or Celsius. One obtains a number, but one cannot be sure of its meaning.  Although it could be argued that all of the variations of the VAS correlate highly, it should be pointed out that Fahrenheit and Celsius correlate a perfect 1.0.  Obviously, though, these two measures of temperature cannot be used interchangeably, as they are entirely different metrics. The same is true of the variations of the VAS.

Medical standards dictate that medications should be safe and effective. Similarly, psychometric standards dictate that a standardized test should be valid (that is, measuring what it is supposed to measure), reliable (meaning that you get the same results upon retest), and that it should have norms that serve as clinical benchmarks to which the patient can be compared. Further characteristics of a standardized test include standardized instructions, standardized test materials, and a manual.

For the most part, the VAS and similar pain rating scales used in the field do not have the above features. An informal survey by the authors found enormous variation in how devices such as the VAS are administered. Using the VAS as an example, the first source of variation comes with regard to what the extreme pain anchor term is.  While the lowest point in the scale is usually described as “no pain,” the highest one can be described in several possible ways.  Some of these are as follows:

Obviously, there are a number of ways for describing the worst possible pain.  Some of these are more neutral, while others confound the issue of pain assessment.  For example, “Pain so bad you want to commit suicide” confounds pain with depression.  While all these forms of pain are related, there are obviously significant differences between them.

A second source of variation in the VAS is the instructions.  Some of the possibilities of the instructions are as follows:

Obviously, the instructions focus the patient’s attention on what is being rated, and influence the patient’s response.

The problems with the VAS and similar measures, such as the lack of normative information, have been suggested by others.  For example, it was observed that:

The appropriateness of norms of tests has rarely been considered in the pain literature. In the absence of normative information, the raw score on any test is meaningless. To observe that a patient with a migraine headache scores a 10 on a Visual Analog Scale (VAS) of intensity conveys little or no information. However, if it is known that the average pain severity for 100 migraine headache patients is 5.4 with a standard deviation of 1.0, this information would permit the conclusion that this patient is expressing a very high level of pain relative to other migraine sufferers. (Turk and Melzack, 1992b)

Turk and Melzack go on to state that general medical pain norms are not the solution, as there would be little value to comparing the pain level of one type of patient to a norm group based on patients with a different diagnosis. It was for this reason that 10 reference groups were developed for the BBHI2. By having 10 reference groups, the BBHI 2 is better able to make comparisons to relevant reference groups.

Development of the BBHI 2 Pain Scale

When the BBHI 2 Pain Scale was being developed, an attempt was made to produce a standardized measure.  First of all, a numerical rating approach was decided upon, as this approach has been found by some researchers to be the most commonly used measure of pain reports (Price, Bush, Long & Harkins, 1994), is easy to administer, is understandable to the patient, provides ratio data, and may avoid rating difficulties inherent in Visual Analog Scale approaches (Jensen, Karoly, & Braver, 1986; Briggs and Closs, 1999).

An 11 point 0-10 scael was chosen as it has a level of sensitivity that is equal to or greater than other approaches (Breivik, Bjornsson and Skovlund, 2000), and provides a sufficient number of levels for discrimination (Jensen, Turner and Romano, 1994). Further, as research has shown that there is little or no gain in reliability in Likert measures with over 7 levels (Cicchetti, Showalter & Tyrer, 1985), there seemed to be little to be gained by utilizing other numerical approaches, such as the 0-100 scaling method.

Secondly, what anchor terms to use was considered. There are considerable differences in the field regarding how to describe the highest level of pain. The authors are aware of clinical settings where pain ratings employ such high-end anchor terms as “Pain so bad that you would want to die.” However, such descriptors would seem to confound pain with depression. One research study compared various anchor terms, and concluded that the the best high end term was “worst pain imaginable” (Seymour, et al, 1985) Consequently, on the Pain Complaints scale, this high end definition of pain was accepted, while the 0 level was defined as “No pain at all.” a phrase which is almost universally accepted.  

Third, instead of viewing pain as a single dimension, the patient was asked to rate pain in 10 areas of the body, and was further asked to respond to four additional pain questions. This approach was derived from clinical tools such as the Ransford Pain Drawing, where diffuse and nonanatomical distributions of body pain were regarded as indicators of psychosocial contributions to back pain (Ransford, Cairns and Mooney, 1976). 

The ten items of the BBHI 2 Pain Complaints Scale can be understood as an attempt to create an objective measure that built on the concepts introduced by the Ransford Pain Drawing projective measure. The Pain Complaints overall score is an index of diffuse, severe pain, which is unlikely to be caused by any single injury or condition. Thus, while there are conceptual similarities to what is being assessed by the two measures, the results of the Pain Complaints scale are more mathematically quantifiable, and are based on national norms. At validation, the Pain Complaints scale was compared with a pain drawing measure. As the original Ransford scoring procedure was intended solely for back pain patients, an alternate procedure was used that awarded points for each area of the body that was pain affected, and which was based on methods that had been well researched (Toomey, Gover and Jones, 1983; Margolis, Tait and Krouse, 1986). This approach was characterized by Jensen and Karoly (1992) as being the most common means of scoring pain drawings. This producing an overall numerical pain score that was found to correlate .70 with the BBHI 2 Pain Complaints scale.

On the BBHI 2’s 10 pain items, the patient was asked to rate the level of pain in the last month.  The rationale for this was as follows. First of all, it was decided that longer time intervals were preferable to shorter ones.  Momentary pain (the pain experienced at the moment one is asked) is less meaningful, as it samples only the moment in time when the patient was filling out the questionnaire.  This may have little to do with how the patient normally feels. Furthermore, momentary pain is likely to have a much lower test-retest reliability, as pain typically varies over the course of a day. Consideration was then given as to whether the interval should be one week or one month. As most outcome studies perform follow-up studies using months as units, such as one month, three month, and 12 month follow-ups, encouraging the patient to consider the level of pain over the last month was consistent with this kind of approach. 

The BBHI 2 Pain Complaints Scale is a double normed scale, standardized on two national samples (see Appendix Double norm interp).  One was a patient sample, and the other a community sample, and both of these samples were stratified to meet U.S. census data for gender, ethnicity, age and level of education. The patient data was further subdivided into a chronic pain group, who comprised almost half of the sample, with the remainder being acute pain patients.  Furthermore, the patients were also subdivided into diagnostic groups by their treating professionals. The groups were headache/head injury patients, neck pain, back pain, upper extremity pain, and lower extremity pain patients.  This grouping allowed for a better comparing of “apples to apples,” as the patient could be compared to another group of patients with a similar diagnosis.

Clinical Interpretation of the Pain Complaints Scale

The BBHI 2 pain scale can be interpreted clinically in a manner similar to pain drawing. For any given medical condition, there is a particular distribution of pain symptoms that would be likely to occur. In a pain drawing, a diffuse pattern of pain symptoms, which is not anatomical in distribution, suggests that psychological factors may be involved in the pain reports. Similarly while a particular pattern of pain reports on the BBHI 2 may be consistent with a physical injury, a diffuse pattern of pain complaints is more suggestive of the presence of psychological factors. The interpretation of the BBHI 2 pain scale differs from pain drawings in some important respects though.  While both can be interpreted clinically, determining by inspection whether or not a patient’s pain complaints are consistent with objective medical findings, the interpretation of projective pain drawings are based largely on clinical intuition. In contrast,  the interpretation of the BBHI 2 pattern of pain complaints is facilitated by the empirical sampling of various types of patients with pain.

On the BBHI 2, the clinician has the option of selecting one of five pain diagnostic categories. Selecting one of these diagnostic categories causes the BBHI 2 report to use the pain reports of a corresponding reference group as a means of comparison. These five reference groups are headache/head injury, neck injury, upper extremity injuries, back injury, and lower extremity injury. For example, if the pain diagnostic category of "back pain" is selected, a second column of numbers will appear on the pain interpretation section of the BBHI 2 report.  The first column of numbers will contain the patient's own pain reports in 10 body areas, while the second column of numbers will display the average pain reports of a national sample of patients with back pain.  The BBHI 2 pain scale thus lends itself to a more empirically based interpretation of pain reports.  This helps the clinician to make more acccurate decisions about whether not a particular patient's pain reports are empirically unusual, or if they are consistent with objective findings.  The patient's pattern of of pain reports on the BBHI 2 is actually able to accurately predict the patient's medical diagnostic category, which is something that has never been demonstrated with pain drawings. This is discussed in detail in the next section.

If none of these five diagnostic categories are selected, the patient's results are instead compared to the pain report of the average person in the community. At first glance, it might be thought that the default comparison should be to compare the patient to the pain scores of the average patient.  However, upon closer inspection, it becomes apparent that this comparison is not the most useful.  The distribution of pain in the average patient is an abstraction, like the average family. One could say that the average family may have 2.7 children, but the reality there are no families with 2.7 children.  Similarly,  as the national patient sample consisted of 40% back injury, and 30% carpal tunnel, the average patient in an abstraction, who is 40% back injured, etc. To say that a particular individual has more neck pain than the average patient is a confusing comparison, and very difficult to interpret.

In contrast, there is more value in comparing a particular patient's reports of pain to that of the average person the community.  The average person the community is a more homogeneous group – it is what normal people experience.  It should be recalled that normal people are not totally pain free, but do typically have some minor pain complaints.  By comparing a particular patient's pain complaints to that of a normal persons, it is possible to assess the degree of abnormality.  It is also possible to assess the degree to which patients' pain complaints have impacted the individuals quality of life, and changed it from that which is normal.

There are further advantages to a multidimensional approach to pain assessment in the clinical setting. Suppose that a patient has a herniated lumbar disc, and reports a low back pain level of 8/10, and leg pain of 8/10. Following a surgical fusion, suppose that back pain is still 8/10, but now the patient reports the leg pain has dropped to 0/10. If the patient is reporting only the overall pain on a VAS, then it would appear that there was no change with surgery, as the peak pain was still an 8 both before and after the surgery. However, to say that nothing happened obviously overlooks some very dramatic effects. Thus, beyond the overall Pain Complaints scale score, the 10 individual pain ratings also have clinical utility, and are listed individually on the BBHI 2 printout.

Mathematical Interpretation of the Pain Complaints Scale

An attempt was made to see how well the 10 BHI Pain Complaints items could predict a diagnosis made by the treating professionals. Using a multivariate statistical approach, discriminant analysis, equations were generated to predict which of five diagnostic categories the patient was most likely to be assigned by the treating professional (headache/head injury, neck pain, back pain, upper extremity pain and lower extremity pain). It was found that the 10 localized pain reports could correctly classify patients diagnostically 82% of the time at cross validation.  

The purpose of this is not to use the BBHI 2 for medical diagnosis.  Instead, the purpose of this is to provide a basis of comparison with a working medical diagnosis.  If the BBHI 2 prediction agrees with the medical diagnostic category, then the medical diagnosis is supported.  However, if the distribution of pain symptoms on the 10 Pain Complaints items is empirically uncommon for an individual with a given medical diagnosis, and if BBHI 2 does predict a different type of injury, it might be worthwhile to reexamine the working medical diagnosis.

There are a number of reasons why the Pain Complaints scale would predict differently than the medical professional. First, it is possible that the patient has an unusual medical condition, or multiple injuries. This would explain the unusual pattern of symptoms. It is also possible that either the caregiver or the BBHI 2 has erred. However, when the pattern of pain reported on the Pain Complaints scale differs from what is normally observed in persons with similar diagnoses, there is an increased risk that psychosocial factors are shaping the pain complaints.

Thus, this mathematical procedure can be seen as having evolved from a Ransford type approach. However, while the Ransford Pain Drawing is a projective technique for back pain patients that involves more subjectivity in interpretation and that is rarely actually scored, the Pain Complaints scale is a standardized objective measure that can be used be all patients, is more quantifiable, and lends itself readily to multivariate analysis.

Additional Dimensions of Pain

There are a number of other dimensions of pain that are useful to assess. One of these is pain range. In general, pain experience is not static, but instead fluctuates from hour-to-hour, and from day-to-day.  The pain range score is an estimate of how much a patient's pain fluctuates over the course of the last month.  This score is calculated by subtracting the lowest overall pain over the course of the last month (item 12) from the highest reported overall pain over the course of last month (item 11).  In some cases, patients will report that the overall highest level of pain in the last month was equal to the overall lowest level of pain in the last month, thus indicating that the pain is totally invariant. However, the nature of pain is to fluctuate, and the report that pain has been totally invariant over the course of the last month is more likely to suggest that a fixated pain perception is present. This could suggest that pain reports are fixed by a cognitive determination, rather than by vicissitudes of nociception.

Another important concept on the BBHI 2 is that of peak pain. Initially, it was believed that peak pain would always be equal to item 11. However, it was subsequently found that test takers often combined their localized pain scores in unexpected ways. As an example of this, suppose a patient complains of headache pain of 6 on our 10-point scale, and back pain of 4.  If you ask what the overall average level of pain is, some will average the two numbers together, saying that their overall pain is about a 5. Other patients will take the pain complaints of 6 and 4 and ignore the 4, regarding their highest overall pain as being equal to their highest localized pain experience. Finally, another group of patients would add local pains of 6 and 4 together, and find that their overall pain level would be rated as an 8.   As a result, it was decided to find peak pain has the highest pain score reported on the first 11 pain scale items.

Peak pain was believed to be important as it is typically this that brings a patient into treatment.  For example, suppose a patient suffers from severe migraine headaches, that occur at a frequency of twice a month, but that in the remainder of the month the patient is pain free.  The patient is not coming into treatment because of the average pain, and is not taking medication for the average pain.  It is the peak pain that is of concern, and the appropriate focus of clinical attention.

One of the difficulties in pain assessment is that due to pain's subjectivity, people apply numbers to their subjective pain experience in different ways.  This does not preclude using pain scores in a relative fashion, to show if a particular patient's pain is more or less that it was previously. However, the fact that patients may assign numbers to their pain experience in idiographic ways makes it much more difficult to determine how a particular pain level will impact an individual's behavior.  In the attempt to control for this, a measure of pain tolerance was introduced.  On the BBHI 2, the maximum tolerable pain was determined by asking patients the item, “What is the highest level of pain you could tolerate and still work and get on with your life?”  This maximum tolerable pain assessment is used as a bridge, using this patient determination to connect the patient’s inner world of subjective pain experience with the outer world of disability behavior.

The BBHI 2 of pain tolerance index represents an attempt to assess the state of the patient’s pain coping capacity.  The pain tolerance index is calculated by subtracting the maximum tolerable pain from the peak pain.  A positive score this scale indicates that patients have some reserve capacity for coping with pain, and that even their worst pain does not have a disabling effect.  In contrast a negative score suggests that the level of pain experienced at times exceeds that with which the patient is able to cope, and thus becomes disabling.  A score of zero then, indicates that the patient at times experiences a level of pain that equals their coping capacity.

For example, if a patient can tolerate a pain level of 6, and the highest pain in the last month was a 3, the Pain Tolerance score would be +3, indicating that there was some reserve pain coping ability. However, if the patient stated that only a pain level of 2 could be tolerated, and that the highest pain in the last month was a 4, the Pain Tolerance score would be -2, indicating that the level of experienced pain was at times intolerable.   A more extreme condition is present when the patient reports that the overall lowest level pain in the last month exceeds the maximum tolerable pain.  Under such circumstances, it can be seen in the patient is reporting that the pain levels are continuously disabling.  Such a patient is said to be reporting continuously intolerable pain.

Effective treatment generally involves improving the pain tolerance score. This may be accomplished either by lowering the level of experienced pain, or through the use of coping techniques increasing the level of pain that the patient can tolerate.

The BBHI 2 overall pain complaints T-score also generates a pain ranking.  This ranking places patients in categories that ranges from scores that are extremely low to ones that are extremely high. In that if a patient is high using the Community T-score, but is in the average range using the patient T-score, the patient’s classified as having a moderately high score.

The BBHI 2 overall pain complaints T-score also generates a pain ranking, which are based on approximate percentile cutoffs. Pain ranking can be useful for the categorizing patients for research purposes, or for clinical triage.  It should be noted that this ranking utilizes both patient and community scores A rank of extremely low corresponds to the lowest 1% of patient's, a very low rank corresponds to being below the 5th percentile, while a low rank corresponds to being below the 16th percentile (or one standard deviation below the mean).  Similarly, a rank of extremely high approximates the 99th percentile, very high approximates the 95th percentile, while a high rank suggests a score above the 84th percentile (or one standard deviation above the mean), all using patient norms. A moderately high scorer is one that is above the 84th percentile for the community norms, but below the 84th percentile in using the patient norms.  The average rank consisted of the remaining patients.

The BBHI 2 uses one critical item to screen for pain that is associated with secondary gain.  Persons who respond positively to this critical item feel that their pain entitles them to some kind of compensation.  The positive endorsement of this critical item strongly suggests that the patient is reflecting on factors which could provide some incentive for pain reports.  When present, this should be taken into consideration when evaluating the patient’s pain. The BHI-2 adds more items for entitlement and compensation focus.  It also adds the content area of items depicting chronic pain ideation.  This is a cognitive dimension to the chronic pain experience, which is quite different from the pain complaints overall score.  The two were found to correlate only .34.

Finally, the BBHI 2 asks the patient to rate the current pain level. This is the only dimension that the VAS is typically used to assess. Given the simplicity of the VAS and its lack of standardization, it is surprising that in many otherwise elaborate medical research studies, this is the sole outcome measure.


As opposed to coming up with a single numerical value, the BBHI 2 allows for a multi-dimensional assessment of pain. The results of the BBHI 2 pain assessment (using the 14 pain items plus the critical items) provides more than a single score. Available information includes the following:

 The results of the BBHI 2 pain assessment should be interpreted in the context of the overall test findings.  It should be recalled that pain is a complicated psychomedical phenomenon.  Consequently, both its experience and report occur in the context of a variety of psychosocial variables. 

Pain can lead to reactive depression or anxiety.  However, it should be recalled that the reverse can also occur.  That is, the presence of depression or anxiety can serve to inflate pain reports.  This occurs because affective distress may serve to erode the patient's adaptive capacity, reducing the ability to tolerate pain and frustration.  Influencing the pain threshold in this way, affective distress may serve to inflate pain reports.  Additionally, from a purely psychophysiological perspective, affective distress has been theorized to affect pain perception, a phenomena that is generally referred to as a gate control theory (Melzack and Wall,1965).  Sometimes, extreme levels of pain that  are  medically  unexplained  can appear  in  combination  with  unusually  low  levels  of  depression  or  anxiety.  Under such circumstances, the possibility that denied affective stress is being somatized should be considered. 

Another factor to consider is the psychosocial context in which the pain symptoms appear may incentivize the patient to either minimize or magnify pain reports.  In this regard, the results of the Defensiveness scale can be helpful.  For example, there are instances where there is secondary gain for the report of pain, such as in a disability evaluation.  However, in other circumstances, such as in a vocational rehabilitation setting, there may be secondary gain for minimizing pain. Overall, monetary incentives have been shown to significantly impact the report of symptoms (Rohling Binder and, Langhinrichen-Rohling, 1995). Thus, changes in the psychosocial environment may lead to changes in the level of pain reports.

Overall, the BBHI 2 offers a standardized means for the multidimensional assessment of pain. It is grounded in broadly accepted approaches to the assessment of pain, is objectively scored, is interpreted using national norms and multiple reference groups, and can be administered as a separate module in about one minute. Research shows that the Pain Complaints scale has a good test retest reliability, that it correlates highly with other widely used measures of pain, and is capable of accurately predicting the medical diagnostic category. It can be used in combination with other BBHI 2 scales and critical items, which will help to determine the overall psychosocial context in which these pain complaints occur.

You are free to copy this document for educational use or for your personal informational use provided:

1) it not edited or modified in any way;

2) no fee or compensation is charged for these copies and

3) all copyright notices remain attached

 Please read Disclaimer