Abstract
The authors report the performance levels and pass rates of various candidate demographic groups and the effect on performance of delaying taking the certifying examination. They also report on the effect American Osteopathic Association reentry resolutions have on allopathic-trained candidates entering the osteopathic certification process in internal medicine. Included in the study were all candidates for the American Osteopathic Board of Internal Medicine certifying examination for the period between 1986 and 2002. Investigators performed group analysis based on type of residency track leading to board eligibility, as well as on the number of retake candidates, candidates reestablishing board eligibility 6 or more years after completion of residency training, and allopathic-trained candidates.
Results indicate that medicine-track candidates performed better than any other study demographic group, including allopathic-trained candidates. A delay in taking the certifying examination after completion of residency results in lower candidate performance and pass rates. Various AOA reentry resolutions have not been successful in the repatriating of allopathic internal medicine–trained candidates into the certification process. Candidates in larger training programs have similar mean performance levels and pass rates as candidates in smaller programs.
Since 1986, the American Osteopathic Board of Internal Medicine (AOBIM) has followed the performance of groups of candidates taking the certifying examination in internal medicine. Each year, the AOBIM reviews group data to identify factors that may be affecting candidate performance and pass rates.
Marked changes have occurred in the training tracks of first-time candidates and in the size of groups taking the examination. Changes in American Osteopathic Association (AOA) policies allowed candidates who were previously ineligible to take the examination to become board eligible. The board analyzed these data to determine the effect of these policy changes on candidate entry in the examination process. The purpose of this article is to report the performance levels and pass rates of various groups and to report the effect of these policy changes and other factors on the entry of candidates into the certification process.
Methods
Demographic and performance data for this study were based on all candidates who took the internal medicine certifying examination between 1986 and 2002. All candidates admitted to the examination satisfactorily completed one of the required AOA residency-training pathways.
Background
Candidates were stratified into groups: first-time takers, retakers, reestablish takers (candidates out of residency training more than 6 years who reestablished their eligibility), and allopathic group (candidates who completed all of their residency training in an allopathic residency program approved by the AOA).
In addition, investigators analyzed first-time takers according to various training tracks: (1) traditional track— candidates complete 1 year of a traditional internship followed by 3 years of internal-medicine training; (2) medicine track— candidates complete a specialty-medicine internship followed by 2 years of internal-medicine training; (3) subspecialty track— candidates complete a traditional internship, 2 years of internal medicine training followed by 1 or more years of subspecialty training; (4) internal medicine/emergency medicine—candidates complete 1 year of internship followed by 4 years of a combined internal medicine/emergency medicine training; and (5) internal medicine/pediatrics—candidates complete 1 year of internship followed by 4 years of a combined internal medicine/pediatrics residency. A reference group included first-time candidates from the six largest osteopathic training programs.
Scientists analyzed the effect of delay in taking the certifying examination. The pass rate for candidates who took the examination in the same year that residency training was completed was compared with the pass rate of candidates who delayed taking the examination for 1, 2, or 3 years after completing the training requirement.
Cumulative pass rates for first-time takers reveal the eventual pass rates of candidates taking the examination in a particular year. Most candidates in this group took the examination one or more times, but approximately 5% of the group took the examination only once.
The gender of the candidates is included in the report. Gender results from the period between 1986 and 1996 were previously reported. This article compares those results to results for the period between 1997 and 2002.
Examination
The 2-day written examination consisted of between 500 (pre-1993) and 550 items of the multiple-choice, single-best answer, and matching type. Items covered the broad domain of internal medicine from traditional subspecialty tracks with the following categories: cardiovascular (13%), pulmonary (11%), endocrinology (10%), hematology/oncology (14%), renal/electrolytes (12%), infectious disease (9%), allergy (5%), gastroenterology (11%), rheumatology (8%), neurology (5%), and miscellaneous (2%). Other important cross-content areas (eg, dermatology, geriatric medicine, disease prevention, substance abuse, rehabilitation medicine) were also covered. Approximately 56% of the items were ambulatory medicine–related items, 19% were in-hospital items, 10% were critical care–related items, and 15% were items in miscellaneous categories. Percentages varied minimally from year to year but remained relatively constant during the past 10 years.
Data Collection
Data collected for each candidate included a logit measure examination score and pass/fail status. Each of the examinations had a block of items in common with one of the previous examinations. By use of psychometric equating procedures,
1 common items were used to adjust scores such that all scores were scaled to the 1986 examination. Thus, all of the examinations were of the same level of difficulty.
In terms of scoring, the Rasch model converts candidates' raw scores to a logit measure in which each logit unit is a linear measurement of ability from the highest to lowest score. A more detailed description of the scoring method can be found in a previous publication.
2 All analyses are based on the entire population of interest; therefore, tests of statistical significance are not provided.
By means of the modified Angoff method, the pass/fail point for the examination is reset every 5 years by AOBIM members, the last set in the 1996 examination. In interim years, the same standard is maintained by use of the Rasch equating method described earlier. As the standard is independent of the performance of any group of candidates, it is possible for all candidates to pass or fail.
Results
A total of 2760 candidates took the AOBIM certifying examination between 1986 and 2002.
Table 1 indicates trends in candidate groups during the past 17 years.
Table 2 segregates first-time takers by training track chosen to become eligible for examination. There has been a progressive decline in candidates who become eligible for examination by completing a traditional-track internship followed by 3 years of internal medicine training or by completing a traditional-track internship followed by 2 years of internal medicine training and 1 or more years of subspecialty training.
Table 1
Number of All Candidates Who Took the Internal Medicine Certifying Examination Between 1986 and 2002 According to Various Group Demographics
| 1986 | 1987 | 1988 | 1989 | 1990 | 1991 | 1992 | 1993 | 1994 | 1995 | 1996 | 1997 | 1998 | 1999 | 2000 | 2001 | 2002 |
|---|
| First time | 61 | 78 | 89 | 81 | 81 | 71 | 92 | 110 | 114 | 96 | 99 | 91 | 128 | 129 | 139 | 136 | 144 |
| Retakes | 29 | 28 | 45 | 59 | 36 | 48 | 40 | 42 | 46 | 60 | 54 | 55 | 45 | 60 | 49 | 45 | 48 |
| Reestablish | 0 | 0 | 0 | 0 | 31 | 18 | 16 | 17 | 41 | 59 | 14 | 18 | 20 | 14 | 16 | 16 | 10 |
| Total | 90 | 106 | 134 | 140 | 148 | 137 | 148 | 169 | 201 | 215 | 167 | 164 | 193 | 203 | 204 | 197 | 202 |
Table 1
Number of All Candidates Who Took the Internal Medicine Certifying Examination Between 1986 and 2002 According to Various Group Demographics
| 1986 | 1987 | 1988 | 1989 | 1990 | 1991 | 1992 | 1993 | 1994 | 1995 | 1996 | 1997 | 1998 | 1999 | 2000 | 2001 | 2002 |
|---|
| First time | 61 | 78 | 89 | 81 | 81 | 71 | 92 | 110 | 114 | 96 | 99 | 91 | 128 | 129 | 139 | 136 | 144 |
| Retakes | 29 | 28 | 45 | 59 | 36 | 48 | 40 | 42 | 46 | 60 | 54 | 55 | 45 | 60 | 49 | 45 | 48 |
| Reestablish | 0 | 0 | 0 | 0 | 31 | 18 | 16 | 17 | 41 | 59 | 14 | 18 | 20 | 14 | 16 | 16 | 10 |
| Total | 90 | 106 | 134 | 140 | 148 | 137 | 148 | 169 | 201 | 215 | 167 | 164 | 193 | 203 | 204 | 197 | 202 |
×
Table 2
Number of First-time Candidates Who Took the Internal Medicine Certifying Examination Between 1986 and 2002 According to Various Training Track Demographics
| 1986 | 1987 | 1988 | 1989 | 1990 | 1991 | 1992 | 1993 | 1994 | 1995 | 1996 | 1997 | 1998 | 1999 | 2000 | 2001 | 2002 |
|---|
| Traditional track | 24 | 42 | 53 | 44 | 53 | 38 | 51 | 41 | 30 | 19 | 22 | 9 | 25 | 25 | 17 | 15 | 18 |
| Medicine track | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 18 | 60 | 48 | 56 | 57 | 90 | 84 | 103 | 96 | 97 |
| Allopathic trained | 0 | 0 | 0 | 1 | 2 | 6 | 8 | 10 | 16 | 17 | 13 | 15 | 3 | 10 | 8 | 9 | 12 |
| Subspecialty track | 37 | 36 | 36 | 37 | 27 | 28 | 34 | 45 | 16 | 12 | 5 | 6 | 2 | 4 | 2 | 2 | 2 |
| Internal medicine/Emergency medicine track | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 4 | 0 | 3 | 4 | 7 | 7 | 11 | 17 | 15 |
| Internal medicine/Pediatrics track | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 1 | 0 | 0 | 0 |
| Reference group | 22 | 28 | 31 | 28 | 41 | 26 | 32 | 44 | 40 | 34 | 39 | 28 | 54 | 59 | 43 | 43 | 49 |
| Total | 90 | 106 | 134 | 140 | 148 | 137 | 148 | 169 | 201 | 215 | 167 | 164 | 193 | 203 | 204 | 197 | 202 |
Table 2
Number of First-time Candidates Who Took the Internal Medicine Certifying Examination Between 1986 and 2002 According to Various Training Track Demographics
| 1986 | 1987 | 1988 | 1989 | 1990 | 1991 | 1992 | 1993 | 1994 | 1995 | 1996 | 1997 | 1998 | 1999 | 2000 | 2001 | 2002 |
|---|
| Traditional track | 24 | 42 | 53 | 44 | 53 | 38 | 51 | 41 | 30 | 19 | 22 | 9 | 25 | 25 | 17 | 15 | 18 |
| Medicine track | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 18 | 60 | 48 | 56 | 57 | 90 | 84 | 103 | 96 | 97 |
| Allopathic trained | 0 | 0 | 0 | 1 | 2 | 6 | 8 | 10 | 16 | 17 | 13 | 15 | 3 | 10 | 8 | 9 | 12 |
| Subspecialty track | 37 | 36 | 36 | 37 | 27 | 28 | 34 | 45 | 16 | 12 | 5 | 6 | 2 | 4 | 2 | 2 | 2 |
| Internal medicine/Emergency medicine track | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 4 | 0 | 3 | 4 | 7 | 7 | 11 | 17 | 15 |
| Internal medicine/Pediatrics track | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 1 | 0 | 0 | 0 |
| Reference group | 22 | 28 | 31 | 28 | 41 | 26 | 32 | 44 | 40 | 34 | 39 | 28 | 54 | 59 | 43 | 43 | 49 |
| Total | 90 | 106 | 134 | 140 | 148 | 137 | 148 | 169 | 201 | 215 | 167 | 164 | 193 | 203 | 204 | 197 | 202 |
×
The number of candidates with a medicine-track internship progressively increased such that in 2002 they comprised 67% of first-time takers, up from 41% of first-time takers since 1986. Although the number of subspecialty-track candidates has decreased to between 1% and 2% of all candidates during the past few years, the number of trainees entering subspecialty training is much higher. Most candidates in subspecialty training have completed a medicine-track internship and take the certifying examination immediately after completing their internal medicine training.
A review of the number of first-time candidates taking the subspecialty examination between 1998 and 2002 reveals that, on average, 40 candidates take these examinations per year. This indicates that approximately one third of all internal medicine residents enter subspecialty training annually. The number of candidates who completed training in allopathic residency programs reached a peak in 1995 but never comprised a significant percentage of the total.
Candidate Examination Performance
Table 3 summarizes the yearly and cumulative mean equated logit scores.
Table 4 shows the yearly and cumulative pass rates of candidate groups. Minor variations in the mean scores and pass rates of first-time takers have occurred over the years, but of note are the improved scores and pass rates that occurred among first-time takers in the past 3 years.
Table 3
Yearly Mean Equated Logit Scores for All Candidates Who Took the Internal Medicine Certifying Examination According to Various Group Demographics
| 1986 | 1987 | 1988 | 1989 | 1990 | 1991 | 1992 | | | |
|---|
| Overall | 0.51 | 0.41 | 0.27 | 0.32 | 0.26 | 0.25 | 0.29 | | | |
| First-time candidates | 0.60 | 0.42 | 0.39 | 0.43 | 0.34 | 0.41 | 0.39 | | | |
| Repeaters | 0.27 | 0.27 | 0.08 | 0.18 | 0.06 | -0.01 | 0.19 | | | |
| Reestablish | — | — | — | — | -0.04 | 0.06 | -0.08 | | | |
|
1993
|
1994
|
1995
|
1996
|
1997
|
1998
|
1999
|
2000
|
2001
|
2002
|
| Overall | 0.31 | 0.23 | 0.20 | 0.21 | 0.34 | 0.22 | 0.18 | 0.30 | 0.34 | 0.36 |
| First-time candidates | 0.45 | 0.37 | 0.40 | 0.29 | 0.49 | 0.33 | 0.34 | 0.54 | 0.60 | 0.60 |
| First-time medicine track | 0.53 | 0.39 | 0.44 | 0.31 | 0.54 | 0.39 | 0.35 | 0.59 | 0.69 | 0.66 |
| First-time traditional track | 0.39 | 0.28 | 0.48 | 0.37 | 0.29 | 0.23 | 0.46 | 0.39 | 0.13 | 0.34 |
| First-time specialty track | 0.47 | 0.49 | 0.29 | 0.27 | 0.37 | 0.12 | 0.15 | 0.25 | 0.53 | 0.45 |
| Reference group | 0.59 | 0.28 | 0.39 | 0.34 | 0.58 | 0.49 | 0.46 | 0.70 | 0.86 | 0.71 |
| Reestablish | -0.13 | -0.08 | -0.03 | 0.05 | 0.14 | -0.03 | -0.10 | -0.12 | -0.19 | -0.14 |
Table 3
Yearly Mean Equated Logit Scores for All Candidates Who Took the Internal Medicine Certifying Examination According to Various Group Demographics
| 1986 | 1987 | 1988 | 1989 | 1990 | 1991 | 1992 | | | |
|---|
| Overall | 0.51 | 0.41 | 0.27 | 0.32 | 0.26 | 0.25 | 0.29 | | | |
| First-time candidates | 0.60 | 0.42 | 0.39 | 0.43 | 0.34 | 0.41 | 0.39 | | | |
| Repeaters | 0.27 | 0.27 | 0.08 | 0.18 | 0.06 | -0.01 | 0.19 | | | |
| Reestablish | — | — | — | — | -0.04 | 0.06 | -0.08 | | | |
|
1993
|
1994
|
1995
|
1996
|
1997
|
1998
|
1999
|
2000
|
2001
|
2002
|
| Overall | 0.31 | 0.23 | 0.20 | 0.21 | 0.34 | 0.22 | 0.18 | 0.30 | 0.34 | 0.36 |
| First-time candidates | 0.45 | 0.37 | 0.40 | 0.29 | 0.49 | 0.33 | 0.34 | 0.54 | 0.60 | 0.60 |
| First-time medicine track | 0.53 | 0.39 | 0.44 | 0.31 | 0.54 | 0.39 | 0.35 | 0.59 | 0.69 | 0.66 |
| First-time traditional track | 0.39 | 0.28 | 0.48 | 0.37 | 0.29 | 0.23 | 0.46 | 0.39 | 0.13 | 0.34 |
| First-time specialty track | 0.47 | 0.49 | 0.29 | 0.27 | 0.37 | 0.12 | 0.15 | 0.25 | 0.53 | 0.45 |
| Reference group | 0.59 | 0.28 | 0.39 | 0.34 | 0.58 | 0.49 | 0.46 | 0.70 | 0.86 | 0.71 |
| Reestablish | -0.13 | -0.08 | -0.03 | 0.05 | 0.14 | -0.03 | -0.10 | -0.12 | -0.19 | -0.14 |
×
Comparison of the pass rates for candidates of various training tracks reveals that the rates for traditional-track candidates (71%), medicine-track candidates (80%), and subspecialty-track candidates (70%) are similar, with a slightly higher pass rate in the medicine-track group. Candidates out of training for 6 years or more (reestablish group) had pass rates slightly lower (35%) than the retake group (44%). The reference group, composed of candidates from the six largest programs, had pass rates (80%) that are similar but slightly higher than the overall first-time takers group (74%).
Table 5 summarizes the performance and pass rates of candidates who completed internal medicine training in an allopathic program. Of all 136 candidates, 44 failed the American Board of Internal Medicine (ABIM) examination before taking the AOBIM examination. The pass rate for candidates who bypassed the ABIM examination and took only the AOBIM examination was 61%. This compares with the mean pass rate of candidates who passed the ABIM examination and then took the AOBIM examination (95%) and to the mean pass rate of the group that failed the ABIM examination initially and then took the AOBIM examination (50%).
Table 5
Allopathic-Trained Candidates
| Year | n * | Mean Logit | Pass † | Failed ABIM ‡ |
|---|
| 1989 | 1 | 0.05 | 0 | 0/0 |
| 1990 | 2 | 0.57 | 1 | 0/1 |
| 1991 | 6 | 0.52 | 5 | 1/5 |
| 1992 | 8 | 0.06 | 4 | 1/3 |
| 1993 | 10 | 0.24 | 5 | 4/4 |
| 1994 | 14 | 0.02 | 5 | 11/12 |
| 1995 | 21 | 0.25 | 11 | 15/15 |
| 1996 | 17 | 0.12 | 7 | 2/7 |
| 1997 | 15 | 0.44 | 10 | 1/15 |
| 1998 | 3 | 0.31 | 2 | 0/0 |
| 1999 | 10 | 0.12 | 7 | 1/1 |
| 2000 | 8 | 0.78 | 7 | 1/4 |
| 2001 | 9 | 0.38 | 9 | 4/5 |
| 2002 | 12 | 0.64 | 11 | 2/6 |
| Total | 124 | 0.27 | 73/124 | 41/73 |
Table 5
Allopathic-Trained Candidates
| Year | n * | Mean Logit | Pass † | Failed ABIM ‡ |
|---|
| 1989 | 1 | 0.05 | 0 | 0/0 |
| 1990 | 2 | 0.57 | 1 | 0/1 |
| 1991 | 6 | 0.52 | 5 | 1/5 |
| 1992 | 8 | 0.06 | 4 | 1/3 |
| 1993 | 10 | 0.24 | 5 | 4/4 |
| 1994 | 14 | 0.02 | 5 | 11/12 |
| 1995 | 21 | 0.25 | 11 | 15/15 |
| 1996 | 17 | 0.12 | 7 | 2/7 |
| 1997 | 15 | 0.44 | 10 | 1/15 |
| 1998 | 3 | 0.31 | 2 | 0/0 |
| 1999 | 10 | 0.12 | 7 | 1/1 |
| 2000 | 8 | 0.78 | 7 | 1/4 |
| 2001 | 9 | 0.38 | 9 | 4/5 |
| 2002 | 12 | 0.64 | 11 | 2/6 |
| Total | 124 | 0.27 | 73/124 | 41/73 |
×
Table 6 summarizes the cumulative pass rates of first-time takers who entered the examination process between 1986 and 1997. These data include all first-time takers with a minimum of 6 years of examination eligibility, though only a minority of candidates took the examination each year of their eligibility. As demonstrated in
Table 6, 94.1% of candidates who took the examination are now certified. Except for 1996, cumulative pass rates for each year of first-time takers has exceeded 92%.
Table 6
Yearly and Total Cumulative Pass Rates, %, of All Candidates Who Entered the Internal Medicine Certifying Examination Between 1986 and 1997 and Took the Examination One or More Times
| 1986 | 1987 | 1988 | 1989 | 1990 | 1991 | 1992 | 1993 | 1994 | 1995 | 1996 | 1997 | 1986-1997 |
|---|
| Pass | 57 | 70 | 77 | 76 | 79 | 67 | 89 | 104 | 106 | 91 | 89 | 86 | 991 |
| Pass, % | 95 | 92 | 94 | 94 | 99 | 94 | 97 | 95 | 93 | 96 | 90 | 95 | 94.1 |
| Total first time | 60 | 76 | 84 | 81 | 80 | 71 | 92 | 110 | 114 | 95 | 99 | 91 | 1053 |
Table 6
Yearly and Total Cumulative Pass Rates, %, of All Candidates Who Entered the Internal Medicine Certifying Examination Between 1986 and 1997 and Took the Examination One or More Times
| 1986 | 1987 | 1988 | 1989 | 1990 | 1991 | 1992 | 1993 | 1994 | 1995 | 1996 | 1997 | 1986-1997 |
|---|
| Pass | 57 | 70 | 77 | 76 | 79 | 67 | 89 | 104 | 106 | 91 | 89 | 86 | 991 |
| Pass, % | 95 | 92 | 94 | 94 | 99 | 94 | 97 | 95 | 93 | 96 | 90 | 95 | 94.1 |
| Total first time | 60 | 76 | 84 | 81 | 80 | 71 | 92 | 110 | 114 | 95 | 99 | 91 | 1053 |
×
Table 7 summarizes the effect on candidates' pass rates of delaying taking the certifying examination; any delay in taking the examination results in a lower pass rate. Candidates who take the examination in the first year of eligibility following completion of residency training have a mean pass rate of 81%, which is higher than the mean pass rates of candidates who delay taking the examination for 1 year (62%), 2 years (44%), or 3 years (50%).
Table 7
Yearly and Total Pass Rates for First-Time Takers Grouped According to Whether They Took the Examination the First Year of Eligibility (No Interval) or Whether There was a Delay in Taking the Examination After Completing Residency Training (1- to 3-Year Delay)
| 1993 | 1994 | 1995 | 1996 | 1997 | 1998 | 1999 | 2000 | 2001 | 2002 | Total | % Pass |
|---|
| No interval | 62/80 | 52/74 | 55/76 | 51/75 | 60/78 | 79/106 | 89/106 | 119/130 | 105/120 | 111/120 | 783/965 | 81% |
| 1-Year delay | 7/14 | 17/26 | 6/12 | 3/8 | 4/8 | 7/12 | 11/15 | 5/8 | 8/10 | 8/9 | 76/122 | 62% |
| 2-Year delay | 4/8 | 2/2 | 4/7 | 1/1 | 2/4 | 1/6 | 1/5 | 0/0 | 1/3 | 0/0 | 16/36 | 44% |
| 3-Year delay | 3/5 | 1/3 | 0/0 | 1/2 | 0/1 | 2/4 | 2/3 | 0/0 | 1/1 | 1/3 | 11/22 | 50% |
| 1- to 3-Year delay | 14/27 | 20/31 | 10/19 | 5/11 | 6/13 | 10/22 | 14/23 | 5/8 | 10/14 | 9/122 | 103/180 | 57% |
Table 7
Yearly and Total Pass Rates for First-Time Takers Grouped According to Whether They Took the Examination the First Year of Eligibility (No Interval) or Whether There was a Delay in Taking the Examination After Completing Residency Training (1- to 3-Year Delay)
| 1993 | 1994 | 1995 | 1996 | 1997 | 1998 | 1999 | 2000 | 2001 | 2002 | Total | % Pass |
|---|
| No interval | 62/80 | 52/74 | 55/76 | 51/75 | 60/78 | 79/106 | 89/106 | 119/130 | 105/120 | 111/120 | 783/965 | 81% |
| 1-Year delay | 7/14 | 17/26 | 6/12 | 3/8 | 4/8 | 7/12 | 11/15 | 5/8 | 8/10 | 8/9 | 76/122 | 62% |
| 2-Year delay | 4/8 | 2/2 | 4/7 | 1/1 | 2/4 | 1/6 | 1/5 | 0/0 | 1/3 | 0/0 | 16/36 | 44% |
| 3-Year delay | 3/5 | 1/3 | 0/0 | 1/2 | 0/1 | 2/4 | 2/3 | 0/0 | 1/1 | 1/3 | 11/22 | 50% |
| 1- to 3-Year delay | 14/27 | 20/31 | 10/19 | 5/11 | 6/13 | 10/22 | 14/23 | 5/8 | 10/14 | 9/122 | 103/180 | 57% |
×
Since 1997, the percentage of first-time female takers has increased from 22% during the period between 1986 and 1996 to 36% during the period between 1997 and 2002. In 1997, it was reported that the mean scores and pass rates were nearly identical in both gender groups.4 Those results have remained unchanged since 1997.
Discussion
In 1989, the AOA approved the establishment of specialty-internship tracks, which reduced the number of required years of internal medicine residency training from 3 to 2 years. As a result, a marked change in the demographics of the candidate pool taking the certifying examination occurred.
Before 1992, approximately 56% of candidates completed a traditional internship followed by 3 years of internal medicine training. The remainder of candidates enrolled in subspecialty training in which the first year applied toward the third year of internal medicine training. A national trend in manpower needs publicized by the federal government followed in which residency programs were encouraged to produce more general internists because there was an oversupply of subspecialists.
The change in the percentage of osteopathic internal medicine residents entering internal medicine practice during the past 5 years can be closely approximated by comparing the number of first-time takers in subspecialty examinations to the total number of first-time takers for the internal medicine certifying examination during an equivalent period. It is apparent from these data that the percentage of candidates entering general internal medicine has increased from 56% before 1992 to 70% for the period between 1994 and 2000. With development of combined internal medicine/emergency medicine residency programs in 1990, there has been a slow growth of candidates completing this training track. Individual program surveys (personal communication) indicate, however, that most of these candidates select a pure emergency medicine career at least during the initial years of their medical practice.
During the mid-1980s and thereafter, the AOA approved several amnesty or reentry resolutions that established pathways by which osteopathic medical school graduates who completed non–AOA-approved residency programs and non-AOA internships to obtain AOA approval of their training. This would allow osteopathic medical school graduates to become board-eligible by one of the AOA certifying boards.
Data published annually by the American Medical Association and by the AOA during the past 14 years indicate that an equal number or more of osteopathic physicians selecting an internal medicine career have completed allopathic residency programs over osteopathic residencies. During the past 5 years, two thirds of osteopathic medical school graduates selected allopathic programs, while one third selected osteopathic programs. It is evident that AOA policy changes that would open doors to osteopathic graduates to reenter the profession and become AOA-certified have not resulted in any significant movement, at least in the field of internal medicine.
Since 1989, the AOBIM has examined 136 allopathic-trained candidates, which comprise between only 3% and 4% of the total allopathic-trained group in allopathic internal medicine training programs. A review of the demographics of this group reveals that 44 of the 136 allopathic-trained candidates entered the osteopathic certification program as a “second chance” method, as those candidates were previously unable to pass the ABIM certifying examination on one or more tries. If one excludes this group, it is apparent that between only 2% and 3% of eligible candidates voluntarily selected the osteopathic certification program as their first choice or as an additional certification.
In 1997, it was reported that the mean logit scores and pass rates for specialty-track candidates and candidates with a traditional internship and 3 years of internal medicine training were identical.
3 Since that time (1997–2002), though overall numbers are small, medicine-track candidates have had higher mean scores and slightly higher pass rates (84% vs 74%) (
Table 8).
Table 8
Comparison of Pass Rates for Various Groups 1986–1996 and 1997–2002
| 1986-1996 | | 1997-2002 | |
|---|
| n | Pass Rate, % | n | Pass Rate, % |
|---|
| First time | 972 | 70 | 767 | 81 |
| Traditional track | 417 | 70 | 109 | 74 |
| Medicine track | 182 | 70 | 527 | 84 |
Table 8
Comparison of Pass Rates for Various Groups 1986–1996 and 1997–2002
| 1986-1996 | | 1997-2002 | |
|---|
| n | Pass Rate, % | n | Pass Rate, % |
|---|
| First time | 972 | 70 | 767 | 81 |
| Traditional track | 417 | 70 | 109 | 74 |
| Medicine track | 182 | 70 | 527 | 84 |
×
When the medicine-track internship was instituted in 1989, there was concern by the American College of Osteopathic Internists Council on Education and Evaluation that shortening the internal medicine residency would result in a lesser fund of knowledge and clinical skills of candidates in the medicine-track program. During the past 9 years, however, medicine-track candidates have performed as well or better than those with 1 additional year of residency training. Since 1993, medicine-track candidates' pass rates have been significantly higher (80%) than the pass rates of traditional-track candidates (71%).
It is evident from
Table 7 that candidates who take the certifying examination when first eligible perform better than those who delay taking the examination 1 or more years. One can assume from these data that candidates' fund of knowledge may diminish as the length of time increases from the end of residency training to taking the examination. However, other factors may play a role in the lesser performance of candidates who delay. These individuals may not represent a cross section of those who take the examination when first eligible. Those who delay may represent candidates who do not feel prepared for examination for various reasons and thus delay taking the examination at the first opportunity. It appears from these data that program directors should encourage residents to prepare well for the certifying examination and take the examination on completion of residency.
Preliminary 1997 data revealed that the performance of allopathic-trained candidates taking the AOBIM certifying examination was lower than the performance of candidates in an osteopathic residency program.
3 At that time (1989–1996), only 50 candidates had taken the AOBIM examination; however, that trend in performance has continued through the 2002 examination. Of 70 candidates who elected to take the AOBIM examination instead of the ABIM examination, the average pass rate was 61%, which is nearly identical to the allopathic-trained group (62%). There were 21 candidates who took the ABIM examination initially and passed and then took the AOBIM examination. Of the 21 candidates, 20 passed the AOBIM certifying examination. Were these 21 candidates removed from the total group of 136, the pass rate for the remaining 115 candidates would be 56%.
As surmised from 1997 data, allopathic-trained candidates taking the AOBIM examination may not be a representative sample of the entire group of residents that trained in allopathic programs. However, the overall pass rates of osteopathic medical school graduates taking the ABIM examination are similar to those taking the AOBIM examination. The question remains whether osteopathic programs are able to attract a slightly better performing group of candidates into their programs, though it must be emphasized that the numbers are small, and any conclusions must be made with caution.
Wright BD, Linacre JM. Bigsteps: Rasch-model Computer Program. Chicago, Ill: MESA Psychometric Laboratory, University of Chicago, 1991.
Slick GL, Dolan S, Draba RE. Performance of osteopathic medical school graduates on the American Osteopathic Board of Internal Medicine certifying examination 1985 to 1994. J Am Osteopath Assoc. 1995; 95:370-374.
Slick GL. Comparison of performance on the American Osteopathic Board of Internal Medicine certifying examination 1986 to 1996 by type of residency. J Am Osteopath Assoc. 1997;97 : 417-422.