Telehealth for Terminally Ill Patient What is Telehealth
Telehealth is defined as the use of interactive technology to provide clinical health care, patient and professional education, and health care administration across small and large distances (American Nurses Association, 1999; Chaffee, 1999). The essential feature of telehealth is using electronic signals to transfer different types of information from one site to another, which may include clinical records, health promotion instructions, still images of wounds, and videos demonstrating exercise routines.
In health sciences literature, telehealth is often used interchangeably with telemedicine, and occasionally, the term telenursing appears. However, telehealth is considered a more encompassing concept, representing the state of technology used in health care delivery, while telemedicine and telenursing are subsets of telehealth.
Telehealth and Nursing
Telehealth offers significant potential for nursing as a means of communication between nurses, patients, and caregivers, and as a way to deliver customized nursing services. Telehealth can be applied across nearly all areas of nursing care, from emergency response systems to hospital, home, and community care settings.
Telehealth expands healthcare services beyond traditional geographic boundaries, providing access to a broader range of care options in underserved areas and during times when healthcare providers are typically unavailable. It can be utilized for bedside nursing care, patient education, and assisting nursing care in remote locations.
This broad definition of telehealth includes several modes of transmission, such as telephone and fax, interactive video and audio, store-and-forward technology, patient monitoring equipment, electronic patient records, electronic libraries and databases, the Internet and intranet, the World Wide Web (WWW), email systems, decision and care planning support systems, and electronic documentation systems.
When effectively employed, telehealth can leverage limited healthcare resources to better meet patient needs (Darkins, Fisk, et al., 1996; Wakefield, Flanagan, et al., 2001).
Most nurses are already participating in telehealth without realizing it, through activities like telephoning or faxing patient status reports, telephone triage, remote home health visits for monitoring, and creating websites for patient education.
Although much focus has been on technology and innovative equipment to improve access and availability of healthcare services for patients regardless of location, there has been limited examination of the efficiency and effectiveness of these applications. An exception is telephone consultations; randomized clinical trials have demonstrated the efficacy of telephone consultations in improving patient outcomes across diverse populations (Balas, Jaffrey, et al., 1997).
Studies of interactive “teleconsultations” worldwide generally suggest that healthcare delivery via these technologies is acceptable to patients in a variety of circumstances.
Telenursing and Benefits
Telenursing offers additional possibilities for nursing practice. A number of research projects (Whitten, Mair, et al., 1997; Hayes, Duffey, et al., 1998; Wootton, Loane, et al., 1998; Hagan, Morin, et al., 2000; Hanson, EJ, & Clarke, 2000; Hanson, EJ, Tetley, et al., 2000; Johnson-Mekota, Maas, et al., 2001; Jerant, Azari, et al., 2003) have identified key components of home care that could be delivered through telecommunications.
One advantage of using Structural Equation Modeling (SEM) to estimate models containing causal paths among latent variables is that many of the assumptions of regression can be relaxed or estimated. For example, unlike multiple regression, which assumes perfect measurement (i.e., no measurement error), SEM can specify and estimate measurement errors.
Additionally, constraints can be introduced based on theoretical expectations. For example, equality constraints, which set two or more paths to have equal values, are useful when the model contains cross-lagged paths from three or more time points. These constraints can help compare models for different groups, such as the effects of maternal employment on preterm versus full-term child outcomes.
Data Requirement
The data requirements for SEM are similar to those for factor analysis and multiple regression regarding the level of measurement, but differ concerning sample size. Exogenous variables can have indicators measured at the interval, near-interval, or categorical levels, while endogenous variables must have indicators measured at the interval or near-interval levels.
A common rule of thumb for the number of cases needed for SEM is 5 to 10 cases per parameter to be estimated, suggesting that considerably larger samples are often required compared to multiple regression, ranging from 100 for modest models to 500 or more for complex models. Despite the advantages of SEM, these larger samples can result in more complex and costly studies.
Structural Equation Modeling is generally a multistage procedure. First, the SEM implied by the theoretical model is tested, and the model’s fit to the observed data is evaluated. A nonsignificant chi-square (χ²) indicates an acceptable fit, but this is difficult to obtain as the χ² value is heavily influenced by larger sample sizes. Therefore, most analytical programs provide alternative measures of fit, which are necessary before evaluating and interpreting parameter estimates.
Modification to SEM
Often, the original theoretical model does not fit the data well, and modifications are necessary to achieve a good-fitting model. Although nonsignificant paths can be deleted, modifications generally focus on adding omitted paths (causal or correlational). Paths that are omitted specify no relationship, implying a parameter of zero; therefore, analysis programs constrain these paths to be zero.
After estimating the specified model, most programs provide an estimate of the “strain” experienced by fixing parameters to zero or the improvement in fit resulting from freeing these parameters (allowing them to vary). Suggested paths must be theoretically justifiable before they are added to the revised model.
Model Specifications
The model specification is based on the data at hand and theoretical evidence, with repeated testing of those data, resulting in a significance level that is actually higher than the program indicates. Thus, additional criteria are necessary to evaluate the final model’s adequacy.
Firstly, the final model must be theoretically appropriate. Comparing the original model with the final model reveals how much “trimming” has taken place. The values and signs of the parameters are also evaluated. The parameter signs (positive or negative) should align with the expected direction.
Parameters on the paths between latent variables and their indicators should be greater than 0.50 but less than or equal to 1.0 in a standardized solution. The lower the unexplained variance of the endogenous variables, the better the model performs in explaining those variables, similar to the 1-R value in multiple regression.
Results consistent with prior expectations and findings from previous research increase confidence in the model and its applicability to disease treatment. For healthcare providers, especially nurses, feeding and hydration of patients, including subcutaneous fluids, were considered appropriate if it comforted family members (McAulay, 2001).
Telehealth as Family Need
Research has shown that the needs of families of terminally ill patients are a critical area of concern. Caregivers often face increasing health issues and struggle to get doctors to take their concerns about symptoms seriously, adding stress to both the patient and caregivers (Davis, BD, Cowley, & Ryland, 1996).
Interestingly, Tang, Aaronson, and Forbes (2004) found that terminally ill patients who did not live with their caregivers had more social support, less pain intensity, higher spirituality, and a significantly better quality of life.
Telehealth as Caregiver Need for Terminally Ill Patients
Caregivers of terminally ill children face unique challenges. Factors influencing how families navigate caregiving include their relationship with healthcare providers, the availability of information, and effective communication between parents (Steele, RG, 2002).
A study of a nurse-facilitated intervention with 24 family caregivers found that information, education, encouragement, and support were crucial needs (Mok, Chan, Chan, & Yeung, 2002). The importance of a trusting relationship with caregivers and confidence that the caregiver will not be abandoned was also found to be essential, reinforcing earlier findings.
A study in Great Britain examining the adequacy of primary care and the deployment of visiting nurses underscored the need for information and support for caregivers (Beaver, Luker, & Woods, 2000). In examining the management of time for caregivers of a terminally ill person, KE Rose (1998) observed that uncertainty about how long they would have with the patient created additional stress.
Despite the uncertainty, many caregivers find meaning in their caregiving roles, helping them develop a new perspective on life and reach out to support others.
In another qualitative study, Duke (1998) explored the experience of caring for a terminally ill spouse and then grieving afterward. Themes that emerged included sharing the illness experience, providing care and comfort, existing in a state of limbo, and recognizing the creation of cherished memories.
Research repeatedly emphasizes the importance of information for caregivers of terminally ill individuals. In Norway and Sweden, researchers found that respondents supported ongoing disclosure of information to terminally ill patients (Lorensen, Davis, Konishi, & Bunch, 2003), which contrasts with practices in parts of Europe and Japan where information is shared with the family rather than the patient. Similar views were found among younger and more affluent Korean Americans and Mexican Americans in a survey of elderly residents in Los Angeles County (Blackhall, Murphy, and Frank, 1995).
Caregivers often serve as the source of information about the end-of-life experience of the terminally ill person. Satisfaction with care is measured by the reports of family members of the patient’s experience. However, research has shown variable agreement with earlier statements made by family members, raising questions about the reliability of such measurements (Hinton, 1996).
Given the challenges of research with terminally ill patients, qualitative research is often favored. However, there is a need for quantitative methods to improve generalizability in future studies. As research findings accumulate, translating them into practice will enhance care for the terminally ill.