Abstract
Background:
Biofeedback is the process by which individuals, using external monitoring devices, learn to control physiologic functioning through psycho-physiologic self-regulation. Biofeedback is reported to have therapeutic effects on conditions such as chronic pain and anxiety, but few studies have investigated the effect on healthcare resource utilization of incorporating biofeedback into the management plan of adolescents/young adults with these complaints.
Materials and Methods:
Fifty-two patients, identified from a hospital database, were seen for new patient visits regarding candidacy for thermal biofeedback-enhanced relaxation sessions by the first author at our adolescent medicine clinic between January 1, 2014, and June 30, 2015. Variables of interest included total number of office visits, phone calls and e-mails (from patients or parents), emergency department visits, and hospitalizations during the year before and after the first biofeedback visit.
Results:
Of the 52 patients, other medical or complementary management was recommended in lieu of biofeedback for seven patients, and they were therefore excluded from the study. Diagnoses of patients accepted for biofeedback included anxiety, headache, gastrointestinal symptoms including abdominal pain, and amplified musculoskeletal pain syndrome. The average number of office visits, and phone calls and e-mails all decreased significantly (P = 0.01 and P = 0.03, respectively), and the number of emergency room visits decreased from 12 to 5 (P = 0.06).
Conclusions:
Thermal biofeedback-enhanced relaxation can be used effectively to reduce patient/family medical/behavioral resource utilization and thus help to decrease healthcare costs.
Introduction
Biofeedback is a process by which individuals, through the use of external physiologic monitoring devices, can learn to control physiologic functioning through psycho-physiologic self-regulation. Individuals are able to observe their body responses to different situations and thoughts and, as a result, are able to train their physiologic reactivity. Among the forms of biofeedback available are electromyography (EMG), electroencephalography (EEG), heart-rate variability (HRV), and skin temperature. 1
The different types of biofeedback have been demonstrated to have therapeutic effects in a variety of medical conditions such as chronic pain, anxiety, urinary incontinence, and tension and migraine headaches. For example, Ebiloguli et al. examined the use of EMG biofeedback in pediatric patients with overactive bladder syndrome (OAB). Biofeedback helped the patients to become aware of the level of relaxation and tone of their pelvic-floor muscles. The study found that, on average, the patients' lower urinary tract symptom score, which compares OAB and other symptoms before and after treatment, decreased from 16.38 before biofeedback therapy to 8.18 after biofeedback therapy. 2,3
The therapeutic use of biofeedback has also been examined in patients with headache. Grazzi et al. investigated the use of EMG biofeedback-enhanced relaxation training in episodic tension-type headaches in children. During biofeedback sessions, patients focused on exercises to relax different muscle groups and received auditory feedback. The study found that the number of headache days per month decreased in females from 16.4 before treatment to 7.1 after treatment. Similarly, in males, the number of headache days per month decreased from 11.6 before treatment to 3.2 after treatment. 4
Another study investigated the impact of both thermal and EMG biofeedback on reducing anxiety in middle-school students. Students in the biofeedback intervention group participated in thermal biofeedback sessions, as well as EMG biofeedback sessions. These students were found to have significantly lower anxiety scores on the State–Trait Anxiety Inventory, compared to students who did not receive biofeedback training. 5
At the Golisano Children's Hospital Adolescent Medicine Clinic in Rochester, New York, thermal biofeedback-enhanced relaxation sessions are offered as one form of therapy for patients with a variety of chronic pain and anxiety-related conditions. The thermal biofeedback sessions involve monitoring finger temperature, which can be used to assess a state of stress or relaxation. That is, during an acute stress response, blood flow is diverted from non-vital areas such as the gastrointestinal tract, fingers, and toes to the brain, heart, lungs, and large muscles of the arms and legs (“fight or flight”). Because less blood flows to the fingers during a stress response, finger temperature decreases. Conversely, during the relaxation response, there is more blood flow to the fingers, which is reflected by an increase in finger temperature. Thus, patients are able to observe their stress and relaxation responses through a thermal biofeedback-monitoring device on the basis of finger temperature. 6
Although biofeedback has been shown to have therapeutic effects in many different medical conditions, little is known about the effect of using biofeedback as part of the management plan on medical/behavioral resource utilization. Examining this issue was the goal of this study. The primary outcome variables were number of office visits, phone calls and e-mails, emergency department (ED) visits, and hospital admissions before and after the first biofeedback session.
Materials and Methods
Intervention
The typical patient in this study was referred for biofeedback to help manage chronic headache, abdominal pain, or joint or muscle pain that was interfering with social or academic functioning. Most patients also reported anxiety, especially about how to cope with escalating pain.
As part of general history taking, patients were asked if they were ever able to keep low levels of pain from increasing to a level where they had to take medication, ask a parent to provide comfort, or go to the school nurse to be released early to go home. They were then asked to describe what they did to manage these lower levels of pain (e.g., 3–4/10) so that the pain did not progress to an incapacitating level (e.g., 8–10/10). Generally, they reported that they talked to or texted friends, watched TV, played video games, listened to music, or, if they were in school, tried to pay attention to what the teacher was saying. That is, they distracted themselves from their pain by engaging in an activity that focused their attention away from their pain.
During the initial biofeedback training session, blood-flow changes during the stress and relaxation responses, as described above, were discussed with the patient. In addition, the gate control theory of pain perception was added to discussions with patients during the last seven months of the data-collection period. This theory hypothesizes that non-painful input from the environment, such as watching TV or being with friends, closes the “gates” to the central nervous system and inhibits transmission of pain sensation. 7 The gate control theory was presented to explain the neurophysiologic basis for the use of distraction in pain management and to provide the background to congratulate patients about already having a number of strategies to help control their pain successfully.
Following these discussions, the patient participated in a thermal biofeedback session. A temperature sensor, which was connected to a computer, was taped to the patient's finger, and the patient was asked to implement a relaxation technique that he/she had found helpful in the past; deep breathing, thinking about the best vacation ever, or thinking of a song were also suggested as needed. For the particular visual feedback used in the clinic, a sun was displayed that could either remain above the horizon (warmer temperatures) or set below the horizon (cooler temperatures). The patient was asked to try to keep the sun displayed on the computer screen as high above the horizon as he/she could. Thus, it was explained, the sun would rise if the patient's finger temperature increased (feeling more relaxed) and would set if finger temperature decreased (feeling more stressed), allowing the patient to receive real-time feedback. After the biofeedback session was completed (typically 5–7 minutes), a graph of temperature changes was displayed. Thoughts and feelings during the session were discussed to help the patient identify those that were most powerful in inducing relaxation. At the conclusion of the visit, patients were given Biodots®, temperature-sensitive adhesive paper that changes color to reflect skin temperature, to use to practice relaxation techniques outside of computer-based clinic sessions or just to monitor state of relaxation or stress during the day when engaged in regular activities.
It should be noted that at least one family member (usually the mother) or friend was present during the explanation and demonstration sections of the session. This inclusion was deliberate in order to allow the family member or friend to learn about the concepts of the stress response and the neurophysiologic basis for distraction in pain/anxiety management. It also allowed them to see the changes in physiologic reactivity as the patient was able to relax to increase understanding for later discussion and reinforcement at home.
Participants
Fifty-two patients, identified from a hospital database, were seen by the first author for a new patient visit regarding candidacy for thermal biofeedback-enhanced relaxation sessions between January 1, 2014, and June 30, 2015.
Data Collection
Patient charts were accessed and reviewed using the medical center's integrated electronic health record system. The University of Rochester Medical Center Institutional Review Board determined that this study was exempt.
Demographic information collected included patient sex and age in years at the time of the first biofeedback visit. Other variables of interest included the total number of biofeedback visits and the total number of office visits, phone calls and e-mails, ED visits, and hospital admissions before and after the first biofeedback visit. In addition, it was also noted whether the gate control theory of pain perception was discussed with the patient. All data were recorded onto a spreadsheet.
A filter in the electronic record was used to identify only hospital encounters, office visits, telephone calls, and patient e-mails that occurred within the date range of one year before the initial biofeedback visit to one year after the initial visit.
After the encounters were filtered, the description and the encounter note of the initial biofeedback visit were reviewed to determine the reason for the visit as indicated by the visit diagnoses, and to determine whether the gate control theory of pain was discussed with the patient.
Next, the number of office visits, phone calls and e-mails, ED visits, and hospital admissions occurring within the 12 months before and the 12 months following the first biofeedback visit were counted.
Office visits that occurred within one year prior to the first biofeedback visit were sorted into one of three categories based on each visit's description and patient encounter note: (1) visits that directly related to the reason for the biofeedback visit, (2) visits that were indirectly related to the reason for the biofeedback visit, and (3) visits that were not at all related to the biofeedback visits.
Office visits were considered to be directly related to the biofeedback visit if the reason for the visit matched the visit diagnoses of the biofeedback visit. Visits were considered to be indirectly related to the biofeedback visit if the reason for the visit did not match the visit diagnoses of the biofeedback visit but included factors that may have contributed to the reasons for the biofeedback visit. For example, an office visit for colitis was considered to be indirectly related to a biofeedback visit where the biofeedback visit diagnoses were anxiety and headache. Lastly, visits that were considered to be not at all related to the biofeedback visit included acute illnesses and surgical emergencies for which the reason for the visit did not match that of the biofeedback visit. For example, an office visit due to mononucleosis or an acute appendicitis would be classified into this category. For all three of the categories, visits in which the sole purpose of the encounter was laboratory testing were not counted.
Office visits occurring within the year after the first biofeedback visit were categorized and counted in the same manner as the visits prior to the initial feedback visit. It should be noted that the biofeedback visits following the initial biofeedback visit were grouped into a category of their own, as opposed to being included in the category of visits directly relating to the biofeedback visit.
Phone calls and e-mails included all calls and e-mails in which the patient or a patient proxy (usually a parent) received medical advice (e.g., discussion about worsening symptoms or questions about possible changes in care). Phone calls and e-mails were counted and categorized in the same manner as discussed above for the office visits. Calls and e-mails solely regarding laboratory results were excluded from the count. In addition, all e-mails belonging to the same e-mail thread were counted as one e-mail in total.
ED visits were assessed based on the description of the reason for the visit as found in the medical record note and were also categorized and counted in the same manner as described for office visits. Similarly, hospital admissions were assessed based on the discharge summary and were categorized and counted in the same manner as described for the office visits. Patient encounters listed as ED to hospital admission were counted as both an ED visit as well as a hospital admission.
Paired t-tests were conducted to compare resource utilization before and after the initial biofeedback visit.
Results
Fifty-two patients were seen by the first author for a new patient visit regarding candidacy for thermal biofeedback-enhanced relaxation sessions between January 1, 2014, and June 30, 2015. Upon evaluation, seven patients were referred for other medical or complementary management in lieu of biofeedback and were excluded from the study. Alternatives recommended to these patients included acupuncture (for gut motility problems or intractable nausea), participation in school-related activities (for reintegration following a sports injury), pharmacologic management (for gastroesophageal reflux), neurodevelopmental assessment (for probable intellectual impairment), and dietary change (for recently diagnosed lactose intolerance).
Demographics of the 45 study participants are presented in Table 1. The average age of the patients at the time of their first biofeedback visit was 15.4 years. Seventy-three percent of the patients were female.
Table 2 shows the four major categories of diagnoses for the biofeedback visit. Many patients presented with more than one complaint. Anxiety was the most common complaint, occurring in 80% of patients. Headache was the second most common complaint, occurring in 44% of patients.
Numbers reflect ≥1 diagnosis at biofeedback visit.
Patient medical/behavioral resource utilization during the 12 months prior to the first biofeedback visit, as well as during the 12 months following the first biofeedback visit, are summarized in Table 3. Office visits comprised the largest proportion of resource utilization, followed by phone calls and e-mails.
Resource utilization relating to the biofeedback visit diagnoses.
Number of office visits not including biofeedback visits.
The mean number of office visits directly relating to the biofeedback visit diagnoses decreased from 4.2 during the 12 months prior to the first biofeedback visit to 2.78 during the 12 months following the first visit (P = 0.01). Similarly, the average number of phone calls and e-mails decreased from 2.33 to 1.58 (P = 0.03). The total number of ED visits directly relating to the biofeedback visit diagnoses decreased from 12 to 5 after the first visit (P = 0.06). The decrease in the total number of hospital admissions from four before the first biofeedback visit to three after the first visit was not significant (P = 0.27).
The effect of discussing the neurophysiologic basis for the use of distraction in pain management was also examined. Twenty-two patients received an explanation about how distraction can help decrease the severity of pain perception; 23 patients did not receive the explanation, which was not introduced as part of standard management until December 2014. Those who did receive the explanation attended on average 3.1 biofeedback sessions, while those who did not receive the explanation attended on average 1.5 biofeedback sessions (P = 0.002).
Discussion
This study was a retrospective chart review comparing resource utilization before and after an office visit to explain and demonstrate thermal biofeedback-enhanced relaxation training as part of the overall management of pain/anxiety in adolescents. A significant difference was found in the number of office visits as well as phone calls and e-mails relating to the biofeedback visit diagnoses before and after the first biofeedback visit.
Providing information about the neurophysiologic basis for distraction as a method of symptom control appears to have had a significant effect on patient engagement in the biofeedback process. This most likely reflects a better understanding of how pain is perceived and provides an explanation to the patient and family about how self-management through distraction, in particular, can be a major intervention. Physiologic confirmation of the patient's ability to reduce stress and enhance relaxation through deep breathing; thinking of a pleasant place, person, activity, or song; watching TV or playing video games; or engaging with friends is, in itself, empowering and helps patients and their family understand the importance of learning and practicing self-management.
Limitations
In this study, availability of medical records was limited to those in the medical center's electronic record system. It is possible that patients used resources outside of this network, which would affect the count of resources. However, it was assumed that access to these additional resources was the same before and after the biofeedback session. Another limitation in this study is that information relating to the utilization of resources in the Department of Psychiatry was protected by Break-the-Glass and, as such, encounters were assessed based on the encounter's description alone.
Conclusion
The findings from this study demonstrate that biofeedback can be used effectively to help reduce patient/family medical/behavioral resource utilization, including office visits and phone calls/e-mails. Thus, biofeedback appears to represent a promising approach for decreasing healthcare costs.
University of Rochester Medical Center
Golisano Children's Hospital
601 Elmwood Avenue, Box 777
Rochester, NY 14642
Phone: (585) 275-3935
Fax: 585-273-1039
E-mail: oj_sahler@urmc.rochester.edu
Footnotes
Author Disclosure Statement
Neither Olle Jane Z. Sahler, MD, nor Janice Zhao, BA, has any conflict of interest to disclose. ■