Telemedicine has the potential to provide quality cost-effective healthcare in urban and rural populations. The use of synchronous hospital-based telemedicine is well-established for nonsurgical patients. However, there have been no evidence-based studies published on its efficacy for cardiac surgical patients requiring postoperative care. The Michael E DeBakey VA Medical Center (MEDVAMC) in Houston is one of a minority of tertiary care medical centers designated to provide care for patients with complex surgical conditions, such as cardiac disease. In a 4-month retrospective review of MEDVAMC cardiac surgical cases during 2009, 53% of patients lived more than 50 miles away in a geographic area spanning 6 states in the Southeastern US. After surgery, the standard of care is for the patient to return home after being discharged and then return to Houston for a face-to-face (FTF) postoperative clinic visit about 3 weeks later. The overall cost to the VA of arranging for a patient's return includes travel and lodging for both the patient and spouse or caregiver. Such FTF visits usually last about 30 minutes and result in patients being discharged from surgical care or needing only minor additional care. Using telemedicine to evaluate these patients would be expected to 1) save the VA significant travel pay expenses, and 2) expand access to tertiary care at local and rural VA facilities. Many telemedicine programs have been implemented ahead of Class 1 research evidence support. This project was unique in its focus on collecting empirical evidence on provider accuracy when using telemedicine prior to program implementation.
Objectives of the study were to test the accuracy of surgeons' decisions when using clinical video-teleconferencing (CVT) compared to face-to-face (FTF) decisions, to assess provider and patient acceptability of using CVT, to assess the feasibility of using existing VA CVT technology, and to determine the recruitment rate for future studies.
Participants (n = 43) were recruited from Veterans referred to Houston for coronary artery bypass graft or valvular surgeries between November 2010 and March 2011. 100% were male, with Mean age of 64 (SD = 8.3, min = 42, max = 85), and ethnicity of 66% Caucasian, 17% African American, and 17% Hispanic. The study excluded those with a history of aortic dissection/aneurysm procedures or atrial fibrillation diagnosis. Data could not be collected on the first 12 enrollees due to technical barriers. The authors solved these problems by obtaining VA Information Security Office (ISO) approval to use Nefsis Dedicated (www.Nefsis.com) software in place of current VA hardware technology. Once Nefsis software was in place on the VA private network, "Virtual" CVT visits were conducted between provider offices and a clinic exam room at MEDVAMC immediately prior to the patient's first regularly scheduled FTF postoperative visit. Each "Virtual" visit was recorded electronically. Then, two independent attending surgeons reviewed the video recordings and made recommendations on whether the patient should be 1) discharged back to the primary physician, or 2) required to return to postoperative surgery clinic for additional care. Provider decision accuracy was assessed by comparing the 2 reviewers' decisions to the FTF decision. CVT acceptability was assessed using modified versions of the Demiris (2000) questionnaire. Also, qualitative data was collected (n = 3) on the feasibility of using CVT between MEDVAMC and the Conroe clinic. An original sample size (n=26) was calculated for the assessment of provider accuracy and patient and provider acceptability (Phase 1). Out of 45 eligibles, 40 subjects were recruited, with 24 completing the study. Data could not be collected from 12 subjects due to initial technology barriers, 3 either did not return to Houston for their postoperative follow-up, or could not return during CVT program times, and only 1 participant withdrew. A sample size of 10 was calculated to evaluate the technical feasibility of using CVT between Houston and Conroe (Phase 2). Due to technology delays, only 3 were enrolled-all of which completed the study. No non-VA sites were included.
Data analysis showed provider decisions were highly accurate when using CVT (n=23). Overall agreement between the independent reviewers and the FTF decisions was 89%; kappa coefficient (95% C. I.)=0.78 (0.60, 0.96); Specificity=87%; Sensitivity=91%. Providers and patients (n=24) reported high CVT acceptability on the Demiris (2000) 17-Item Questionnaire (range=17 to 85; where 51=acceptability). Provider composite scores (min=58; M=68; max=85; SD=6.7). Patient composite scores (min=49, M=65, max=76, SD=5.6). Existing VA CVT technology was determined to be inadequate for outpatient surgical follow-up care due to lack of availability of HD equipment, lack of equipment with recording compatibility, poor image quality-pixilation, lack of technical support for CVT research, and connectivity problems. However, Nefsis CVT software was found to provide superior image quality, be much more flexible (worked on any desktop workstation on the VA private network that was equipped with a webcam and microphone), and cost significantly less than existing technology in use by the VA ($100 webcam hardware investment for Nefsis vs. $30,000 for a new HD PolyCom set-up). The recruitment rates was high at 90% (43 recruited/48 eligible).
This pilot study provides sufficient data to power a full prospective randomized controlled clinical trial on the use of telemedicine in postoperative cardiac surgical care. Long-term, using telemedicine technology is expected to provide accurate postoperative assessment at lower cost to the VHA system, reduced patient travel effort, and with higher patient satisfaction rates.
- Rodriguez V, Chen GJ, Kunik ME, Pietz KC, Anderson JA, Bakaeen FG, Cornwell LD, Smith DL, Berger DH, Chu D. The use of telemedicine in postoperative cardiac surgical care: A pilot study. Poster session presented at: VA Association of Surgeons Annual Meeting; 2012 Apr 1; Miami Beach, FL.