A total of eighty-nine osteopathic medical students (n = 45), fellows (n = 16), residents (n = 6), and practicing physicians (n = 22) participated in the study. The accuracy of identifying the blinded objects was 82% (SD 17.4%) in Part I, 33% (SD 35%) in Part II, and 0% in Part III for all participants. Overall accuracy for each object is shown in Figure 4.
Accuracy of object identification based on osteopathic level of experience is shown in Figure 5. No statistically significant difference was found in the accuracy of object identification among osteopathic medical students, fellows, residents and practicing physicians.
In this study, accuracy of object identification among all participants was higher in Part I than Part II. This was expected because Part I provided choices, whereas Part II required participants to identify the objects without cues. For Part II, because alternate answers were not determined prior to the study, only absolute answers were accepted. Alternate answers chosen post-testing would have biased result interpretation.
Correct identification was unexpectedly low for some objects. In Part I, 54% of participants accurately identified object ‘a’ as a button and 40% incorrectly identified it as a nickel. Yet when evaluating answers for object ‘e,’ 92% of participants correctly identified the item as a quarter. A possible explanation may be that participants may not have reviewed all fifteen choices available prior to beginning Part I. Thus, participants were not mindful of the subtle difference in the objects. For example, the button was the first item and participants may not have identified the holes and incorrectly answered nickel, thereby eliminating nickel as a future choice. If this is in fact the case, the order of objects and resultant elimination of choices becomes an issue.
Insufficient time dedicated to meaningful touch or lack of reexamination may have also been the cause for low accuracy in object identification. For example in Part III, although 35% correctly identified the item as “hair,” 0% of the participants were able to determine how the hair was arranged. In addition in Part I, about two-thirds of the participants correctly identified object ‘d’ as paper clips, while the remaining incorrectly identified the object as safety pins. Although similar, there are characteristics that help distinguish between these two objects. Irrespective of there being a 10 minute time limit, no participant required additional time for testing. It is possible that the amount of time dedicated to each object varied. This leads to the question of how much time should an examiner palpate? With increased palpatory experience, it is assumed that less time is needed.
While a novel study design, affixing objects against a board created a two-dimensional surface, and the cloth blinding the objects may have further impeded tactile abilities. Three-dimensional testing could have improved accuracy if participants freely handled an object in a blinded container. In this experiment, haptic perception was limited by the cloth and is considered to be indirect palpation because the cutaneous feedback has been limited [14]. In addition, accuracy of object identification may have been poor because despite the objects being “common,” participants may not have had intentional experience in handling these objects. Ability to perform this task could possibly rely more on the ability to visually imagine or mentally represent the object and/or the ability to verbally describe the object being palpated [15]. Familiar objects are more likely to be accurately recognized, but even unfamiliar objects are possibly recognized at levels significantly above chance [16]. Therefore, each palpatory experience of an object should improve accuracy with palpations of that object.
Accuracy of object identification may have improved if a training session was performed prior to testing. Conclusions as to whether the results stemmed from pre-testing training would then need to be addressed. Equal amount of training may result in no difference in accuracy of object identification among different levels of osteopathic experience. However, having clinical palpatory experience may generate a measurable difference because the learning curve may be steeper than for someone with less palpatory experience.
Although this study was successful in obtaining a high number of participants in a short period of time, the population was heavily weighted towards osteopathic medical students. Given the study’s setting this may have been unavoidable, because the American Academy of Osteopathy Annual Convocation is attended more by students than interns and residents. On the other hand, testing at such an event was ideal for attracting practitioners who are actively using osteopathic manipulation and are a better population for measuring a difference in tactile skills.
The finding that no statistically significant difference was found between accuracy of object identification and levels of experience is still surprising. However, measuring clinical palpatory accuracy is more difficult to assess. Osteopathic philosophy and teachings focus on palpation of living and moving tissue, a component of palpation which this study did not undertake. In addition, multiple variables are involved in the evolution of one’s clinical palpatory experience and clinicians depend upon a milieu of clinical skill sets learned from experience that is beyond palpation alone. This study was a simple design, and the complexity of palpatory skills evident among varying levels of experience was not measured and contributed to the lack of perceivable difference. However, future studies on measuring palpatory accuracy as it relates to palpatory experience should be performed. Better designed studies could provide light on how training can lead to improved palpatory accuracy.