Use of Short Message Service and Smartphone Applications in the Management of Surgical Patients: A Systematic Review

Introduction

Surgical outcomes heavily depend on effective patient education and adherence to perioperative protocols. Traditional tools used to educate patients include face-to-face counseling, printed materials, and more recently, tablet-based electronic interfaces. ¹ Similarly, communication between patients and healthcare teams has largely been conducted through telephonic interactions often involving several layers of personnel. Mobile health (m-health) shows tremendous potential for improving communication between patients and providers, as well as increasing patient adherence through reminders and focused education. ^{2 –8}

The application of telemedicine has expanded greatly as cell phone usage has become virtually universal. ⁹ Mobile devices have been incorporated into the medical realm in a variety of contexts ranging from medication reminder applications to algorithms that monitor patient symptoms postoperatively. ^3,10 Numerous studies suggest that short message service (SMS) text messaging or applications that remind the patient to perform health-related activities can markedly increase treatment adherence. ^{2 –8}

The use of telemedicine and m-health in the surgical realm is expanding rapidly. Telemedicine in surgery has been used as an educational tool to inform patients about their procedure and to remind patients to adhere to their prescribed health activities both preoperatively and postoperatively. ^{6,11 –13} SMS has also been implemented for earlier identification of postoperative complications by remotely evaluating patient symptoms after discharge. ^{10,14 –17} Various studies also suggest that telemedicine can reduce healthcare costs by lowering clinic no show rates, readmission rates, and preventable emergency room visits. ^{15 –19}

While individual studies have documented improved perioperative patient care using electronic patient communication systems, there is no broad consensus regarding such applications. Hence, we conducted a systematic review to examine patient-centered outcomes in studies using SMS text messages or mobile applications in the perioperative period in an effort to summarize the data and identify future directions in automated patient communication interventions (APCIs).

Methods

Search process

The literature search was performed in November, 2016 in eight databases using key terms by a medical librarian at Washington University in Saint Louis School of Medicine. The databases include Ovid MEDLINE, Embase, Scopus, the Cumulative Index of Nursing and Allied Health Literature, Cochrane Database of Systematic Reviews along with Cochrane Register of Controlled Trials, ProQuest Dissertations and Theses, and www.Clinicaltrials.gov. In addition, four of the authors also searched PubMed using the same key terms. Key search terms included Surgical Procedures, Operative, Postoperative, Preoperative, Text Messaging, Electronic Mail, and Mobile Applications. The specific search strategy for MEDLINE is shown in Appendix 1.

After removing duplicates, the search resulted in 2,492 unique studies. This initial list of searches was split four ways between four researchers, and studies were selected by applying the inclusion and exclusion criteria to the abstracts. Seventy four (n = 74) studies remained after this preliminary screening. This shorter list of studies was further analyzed using the same inclusion and exclusion criteria by one researcher to identify the 15 studies that are examined in this review. Data were then gathered from these 15 studies by one researcher to standardize the collection process. The PRISMA flowchart details this selection process (Fig. 1).

OPEN IN VIEWER

Fig. 1.

PRISMA flowchart outlining the study selection process.

For each of the studies, we recorded the type of intervention used (SMS or smartphone application), surgical procedure performed, purpose of the messages sent, number of patients enrolled, study design, results, and patient satisfaction with the system, if measured.

Results

After the initial search, a total of 3,048 initial studies were identified. Of these, 525 were duplicates, leaving 2,492 unique studies. A total of 2,451 results were obtained from the keyword literature search and 41 from an independent search in PubMed and MEDLINE. A total of 2,418 articles were eliminated from our review for not meeting the inclusion and exclusion criteria. After further review, another 59 studies were excluded for lacking relevant data, not being a peer-reviewed published article or implementing a purely educational application. All but 15 articles were eliminated. ^{8,13,16 –18,20 –29}

Of the 15 studies that met our criteria, 8 studies utilized SMS, 4 used smartphone applications, 1 used both SMS and a smartphone application, 1 implemented automated phone calls, and 1 study provided phones for pictures to be sent to the physician (Table 1). Of the included studies, 6 were randomized controlled trials (RCTs), 7 were prospective observational, 1 was a retrospective cohort, and 1 was a cross-sectional correlational design study. Six of the 15 studies monitored symptoms, 3 provided patient education (of which 2 promoted postoperative adherence and 1 also provided supportive messages), 3 focused on encouraging patient adherence to medications or protocols, and 3 included a combination of monitoring symptoms while encouraging patient adherence to medications or protocols. The sample sizes ranged from 26 to 1,200. The duration of the interventions spanned from 5 days to 13 months. The studies were performed in 8 countries. None of the studies quantified costs and savings data, but three studies implied fewer clinic or emergency room visits and decreased utilization of resources. ^17,20,29

The studies evaluated involved a variety of procedures (Table 1). Only one study offered a preoperative intervention. ¹³

Discussion

In this systematic review, we evaluated the results of automated SMS and application-based interventions used in the surgical setting. The APCIs evaluated overwhelmingly impacted patient adherence, outcomes, communication, patient satisfaction, and physician workflow positively.

SMS-based interventions were implemented twice as frequently as application-based interventions in the studies evaluated (8 vs. 4, respectively). We hypothesize that smartphone applications are used less frequently because they have higher barriers of entry for the patient than SMS interventions. A smartphone is required to download an application, while any cell phone can send and receive SMS. An SMS-based intervention is preferred over an application-based intervention due to this cost barrier.

In contrast, applications have the potential for a superior graphical interface with more modalities of communication, such as videos and pictures. ^13,29 For example, the Pocket PATH application allows lung transplant patients to track their own progress through charts, facilitating instantaneous feedback. ³¹

m-Health interventions showed promising results in increasing patient adherence. ^{8,22,25,28,31} Patient noncomprehension of discharge instructions has been a documented problem. ³² The hospital is a stressful environment and discharge instructions from clinicians may not be clear to the patient. ³³ Patients may have difficulty understanding and remembering details of discharge instructions. ^34,35 Based on the increased patient adherence suggested by numerous m-health intervention studies, we hypothesize that automated information delivery may be more effective because it is not as overwhelming for the patient. Automated information delivery allows for reminders and feedback from the medical team to patients in the comfort of their home instead of the hospital. These reminders also address unreliable patient recall of discharge instructions. The observer effect, where people behave differently when they are aware of being monitored, may also be involved. This m-health mediated increase in adherence may translate into improved clinical outcomes.

Lin et al. ¹⁸ suggested that SMS-based interventions may have a significant impact on clinic attendance in developing countries, specifically China. Lin et al. ¹⁸ showed a 47.2% increase in ophthalmic clinic attendance post-SMS intervention. While providing medical care to underserved populations in developing countries in a formal clinic may be difficult, cell phone possession is virtually universal. ³⁶ Given the suggested benefits of m-health interventions and high cell phone possession rates, m-health interventions may be an effective tool to monitor and remind patients to perform their health-related activities when they cannot be examined in a formal clinic.

Symptom monitoring by m-health interventions postoperatively reduced the need for patient visits ²⁹ and produced effective alerts for rapid triage and care. ^8,10,20,23 One consideration when determining the threshold for alerts involves whether an alert should be generated for a patient nonresponse to a message. Carrier et al., ²⁰ Miloh et al., ⁸ and Debono et al. ²³ generated alerts for nonresponse. They hypothesized that a nonresponse could have resulted from a negative patient outcome and may necessitate follow-up from the provider. However, the frequency of alerts and how the medical team receives alerts (SMS, e-mail, phone call, etc.) should also be considered. Debono et al. ²³ reported that 19 patients generated 29 alarms. Such a high frequency of alerts requires response from the medical team and is an increased burden.

The advent of fast track recovery programs, such as Enhanced Recovery After Surgery, and a possible movement toward outpatient surgery for specific procedures in fast track units make remote monitoring of systems increasingly important. Length of effective patient monitoring will depend on the procedure and the time window and severity of possible complications, as suggested by the APCIs evaluated. Patient monitoring ranged from 5 days (Carrier et al. ²⁰ ) to over a year (Miloh et al. ⁸ ).

While the relative underuse of mobile phones within geriatric populations may present a barrier to implementation of m-health interventions within this demographic, the studies evaluated in this review suggest that both young and old patients (ages ranging from 2 to 84) can benefit from m-health interventions. ^8,17

Although not directly evaluated, potential cost savings were suggested in some of the studies reviewed. Carrier et al. ²⁰ implied that their intervention resulted in fewer emergency room visits. In Martinez-Ramos et al.'s ¹⁷ study, 55% of patients who reported complications stated that they would have returned to the hospital if not for the intervention. Chen et al. ²¹ suggested that costs and resources could be saved if low-risk patient groups, whose symptoms could be monitored through applications, made fewer clinic visits. Conversely, monitoring of high-risk patient groups and earlier detection of complications may also lead to a reduction in costs. Mobile interventions that reduce the requirement for in-person follow-up may provide a financially advantageous alternative to traditional office visits.

One concern for all m-health interventions is Health Insurance Portability and Accountability Act (HIPAA) compliance and protection of patient health information (PHI). ^37,38 Current standards specify that all m-health interactions should be secured to the “greatest practical extent through use of a private, point-to-point circuit, Integrated Services Digital Network, Advanced Encryption Standard 5 encryption, or virtual private network for Internet transmissions.” ³⁷ Even if the server containing the patient data is secure and the connection is difficult to intercept, record of all interactions will be on that patient's phone. Anyone with access to the patient's phone can observe all transmitted conversations. HIPAA compliant messaging applications exist, but they require use of a smartphone and thus are inaccessible to some patients. ^39,40 Regular SMS cannot be encrypted. ⁴¹ While finding an ideal solution to this issue is difficult, protecting PHI on secure servers, securing the connection, and transmitting the minimal amount of PHI as possible limits patient exposure.

SMS or application usage fatigue, where users may stop responding to the interface, is a limitation for all such interventions. Semple et al. ²⁹ reported that logins were higher in the first 14 days versus days 15 through 30. Miloh et al. ⁸ reported 41% dropout rate by 4 months. Miloh et al., ⁸ Mundi et al., ¹³ and Semple et al. ²⁹ also suggested decreased response rates later into the study. Interventions should be designed in a way to maximize patient retention. An excessive number of notifications, messages, or required responses may decrease engagement. Strategies used in commercial mobile applications may be effective. They include personalization of message content and sending of notifications at times when a user is most likely to view them.

Our review has several limitations. The number of studies included, 15, was relatively small. Only a few articles for each of the categories of monitoring, adherence, and reminders were available to analyze. The studies were heterogeneous, with varying study design, patient demographics, messaging modalities, procedures performed, and outcome measures. This precluded meaningful statistical analysis. Publication bias was also of concern. We attempted to minimize publication bias by searching multiple databases with a wide range of search terms.

Conclusions

This review found that SMS-based and application-based interventions can efficiently and accurately evaluate surgical patient symptoms postoperatively, improve patient adherence to prescribed protocols, improve clinic attendance, and have the potential to reduce healthcare costs. Mobile interventions have high patient satisfaction rates and can be used for wide age ranges. Future issues of concern include the difficulty of maintaining patient engagement and protection of PHI.

Currently, m-health interventions are not widely used, and data reporting their effectiveness are limited. However, m-health interventions have the potential to achieve an amount of precision and personalization in healthcare that present manpower limits cannot support. By utilizing a patient's own technological resources, m-health not only provides useful information to the medical team but also engages the patient in their own healthcare and facilitates joint decision-making between patient and provider. A practical application of these benefits is in addressing patient dissatisfaction with appointment availability. With the information obtained through m-health data, providers can better customize an appointment to target a patient's needs and even avoid unnecessary appointments.