Advances in technology have profoundly influenced the possibilities for clinical assessment. Perhaps the biggest development is a method of assessment that targets people's moods, cognitions, behaviors, physiological states and experiences as they live their daily lives. Ambulatory assessment, which is also referred to as ecological momentary assessment (Stone & Shiffman, 1994), differs from traditional forms of assessment such as questionnaires or clinical and diagnostic interviews in at least two important ways (Trull & Ebner-Priemer, 2013).
First, because ambulatory assessment involves multiple assessments over time, it is uniquely suited to focus on within-individual processes. For example, mood is a dynamic process that ebbs and flows over time, often as a result of contextual or environmental factors. Yet traditional, cross-sectional assessment requires individuals to somehow characterize mood by aggregating across time in some unspecified way. Ambulatory assessment captures slices of this process in real- or near-real time, allowing an evaluation not only of changing mood but also of internal and external influences on mood.
Second, the ecological validity and external validity of ambulatory assessment exceeds that of more traditional measures that are often completed in the artificial environment of the clinic, office or hospital.
Ambulatory assessment: data-collection methods
Before discussing the ethical issues associated with ambulatory assessment, it is important to briefly present some of the data-collection methods that are used. Early research on daily life used paper-and-pencil diaries for participants to record thoughts, feelings and experiences (typically referred to as the experience sampling method). Often, participants were given packets of paper diaries and instructed on when to complete the individual surveys, or they were prompted by pages or timers to complete the surveys. These surveys were then turned in to the researchers at designated intervals or at the end of the study.
The introduction of electronic diaries for recording experiences represented a major advancement for several reasons. First, all entries are time-stamped. Second, there is no need to transcribe or score responses by hand. And third, electronic diaries are more convenient to carry and can store large amounts of data.
Also, and perhaps less familiar, there are "passive" data collection methods for ambulatory assessment research. Behaviors or states can be "observed" by electronic devices. For example, devices can record opening of medication bottles (including digital time stamps), smartphones can record audio or video as well as geographical location, and there are a number of products on the market that can record physical activity levels, cardiac activity, respiration and electro-dermal activity.
The widespread availability and use of smartphones is a game-changer for ambulatory assessment. Almost 60 percent of American adults own smartphones, and there are approximately 2 billion smartphone users worldwide. The smartphone can serve as an electronic diary and can collect data using many of its own built-in sensors and functions. Furthermore, using Bluetooth connections, the smartphone can serve as a wireless hub that collects and transmits data from both internal and external sensors.
Ethical issues and challenges
With this increased utility comes a parallel increase in both ethical issues and assessment challenges. They include:
- Informed consent. As with all forms of assessment, it is necessary to ensure that ambulatory assessment participants are informed about the procedures or protocol of the study, the exact nature of the data to be collected, and potential risks and burdens related to the study. Several unique features of ambulatory assessment should be considered. First, especially because ambulatory assessment may involve passive data collection, it is vital to make the participant aware of all of the data that are being collected, as well as how these data might be used. It is also important to recognize that ambulatory assessment may unintentionally capture data on nonconsenting people who interact with the participant via audio recordings, videos or photos. Investigators must decide ahead of time how this should be handled. Should people be encouraged to discuss their participation in the ambulatory assessment study with others with whom they have contact? Some U.S. states may forbid the recording of third parties without their permission. Participants should be given the option to stop recording at any point and to review their data if recording has occurred in sensitive situations.
-
Confidentiality and privacy. As with any form of assessment, it is important to discuss the limits of confidentiality, such as indications of danger to oneself or to others that will require action on the investigator's part. Equally important is discussing how the investigators plan to ensure privacy in the ambulatory assessment study. Because people are completing ambulatory assessment protocols, including survey questions, in their natural environments, there may be a gap in time between when a participant reveals information that requires some action from the researchers and when researchers actually see the responses. Because of this gap, researchers may choose not to ask for certain information in the study or to ask for this information once the participant is in the laboratory or clinic (such as suicidal thoughts). In some cases, it will be advantageous to seek a certificate of confidentiality from the federal government to protect information regarding illegal activity, such as illegal substance use, from being subpoenaed by third parties, or to protect health information from third parties that may have some incentive to use the information in a way that is disadvantageous to the participant (such as by insurance companies).
In addition, investigators must be mindful about the possibility that study devices such as electronic diaries, wireless sensors and wearable devices may identify an individual as a study participant. Participating in an ambulatory assessment study per se may not be a risk to privacy, but if the study is limited to those who are undergoing mental health treatment or to those with a certain psychiatric diagnosis, then carrying these devices reveals health information. In the end, any risks to privacy or confidentiality need to be revealed and discussed with participants before starting the study. - Data security. More so than in traditional clinical assessment, ambulatory assessment presents challenges in guaranteeing the security of the data. For example, paper diaries may be lost or misplaced, smartphones may be stolen or lost, and wireless transmission of data to servers can be intercepted. Therefore, it is important to take a number of precautions to maintain data security. First, all devices used should be password-protected, and responding to protocols on electronic devices should be password-protected as well. Second, data on the smartphone (whether from answers to surveys or collected from wireless devices) must be encrypted so that if a smartphone is lost, its data cannot be read. Third, in studies that wirelessly transmit data from a smartphone to a server, the data transmission itself must be encrypted. Finally, it is also necessary to encrypt data on the study's secure server to keep unauthorized people from accessing the study data. Investigators must also provide Health Insurance Portability and Accountability Act and institutional review board training to all those that have access to data. In summary, password-protection of the protocol and the devices, as well as several layers of data encryption, are needed to ensure the security of the data in ambulatory assessment studies.
- Assessment issues. Collecting data in the natural environment presents some unique challenges. As mentioned previously, an investigator may collect data that require a response by the investigator (in case of danger to oneself or others, for example), but the investigator may not be made aware of these responses in real time, or may not be accessing the data at certain times of day or evening. One way to address this challenge is to program immediate alerts to inform designated "covering" staff in real time so that interventions or responses can be made in a timely manner. Another issue that may arise involves the collection of biological data. For example, what are the investigator's responsibilities if adverse physiological events or health conditions are detected, such as heart arrhythmias? Researchers should consider these issues before beginning the study, and should put in place protocols to handle such situations. Finally, researchers should adopt protections for privacy and confidentiality, as noted previously, so that study participants are not in jeopardy of losing services to which they are entitled, such as health insurance.
- Treatment and intervention issues. With the widespread adoption of smartphones and tablets, we have seen an explosion of apps purporting to address such mental health issues as depression, anxiety, problematic drinking, problematic eating and more. Although there is great promise in this line of intervention (sometimes called ecological momentary interventions, just-in-time interventions or mHealth interventions), there is little empirical evidence for the efficacy of these interventions administered in daily life (see Kaplan & Stone, 2013). To date, there has been almost no oversight in the development of these daily life interventions from scientific or professional organizations. Furthermore, it is sobering that many of these are not clearly based on principles of behavior change, but rather are developed to look "face efficacious." The question also arises of whether the use of apps for daily life intervention in some way represents an electronic form of the independent practice of psychology. That is, what kind of "supervision" is necessary for these interventions, if any? Clearly, the field of psychology needs to attend to this proliferation of apps that may serve to confuse and mislead those seeking mental health services.
Timothy J. Trull, PhD, is a professor of psychological sciences at the University of Missouri.
