Article Text
PDF
Abstract
Background Technology represents a promising tool to improve healthcare delivery for patients with cirrhosis. We sought to assess utilisation of technology and preferred features of a digital health management tool, in patients with an early readmission for decompensated cirrhosis.
Methods We conducted a cross-sectional study of patients readmitted within 90 days for decompensated cirrhosis. A semistructured interview obtained quantitative and qualitative data through open-ended questions.
Results Of the 50 participants, mean age was 57.6 years and mean (range) model for end stage liver disease was 22.7 (10–46). Thirty-eight (76%) patients own a Smartphone and 62% have regular access to a computer with internet. Thirty-nine (78%) patients would consider using a Smartphone application to manage their cirrhosis. Forty-six (92%) patients report having a principal caregiver, of which 80% own a Smartphone. Patients were interested in a Smartphone application that could communicate with their physician (85%), send medication notifications to the patient (65%) and caregiver (64%), transmit diagnostic results and appointment reminders (82%), educate about liver disease (79%), regularly transmit weight data to the doctor (85% with ascites) and play a game to detect cognitive decline (67% with encephalopathy). Common themes from qualitative data include a desire to learn about liver disease and communicate with providers via digital tools.
Conclusion Among patients with cirrhosis with an early readmission for decompensation, most have Smartphones and would be willing to use a Smartphone to manage their disease. Future digital health management tools should be tailored to the use patterns and preferences of the patients with cirrhosis and their caregivers.
- healthcare innovation
- qualitative research
- telemedicine
- healthcare delivery
- technology
Statistics from Altmetric.com
Introduction
Decompensated cirrhosis represents a substantial source of morbidity for our patients and burden on our healthcare system. Of patients hospitalised with decompensated cirrhosis, 20%–30% are readmitted within 90 days and 50% are readmitted within 6 months.1–7 The estimated national cost of admissions and readmissions for decompensated cirrhosis in 2014 was $1.8 billion and $0.5 billion, respectively.1 As rates of alcoholic and non-alcoholic liver disease rise, the healthcare burden of cirrhosis is expected to grow.8 9
Evidence-based guidelines have made important advances in cirrhosis care; however, many of these guidelines are not sustained in practice.10 11 Reduced guideline compliance has been associated with poor outcomes, including readmissions for overdiuresis, delayed hepatocellular carcinoma (HCC) diagnosis, reduced curative HCC treatment and increased oesophageal variceal haemorrhage.6 10 12 Quality improvement studies of decompensated cirrhosis have demonstrated improved patient outcomes by increasing practice guideline adherence.13–17 An estimated 22% of readmissions for decompensated cirrhosis could have been prevented, in many cases by closer adherence to evidence-based guidelines.6
Technology represents a promising and underused tool to improve patient compliance and effective healthcare delivery to patients with decompensated cirrhosis.16 18 19 Digital health management tools have the potential to reach patients in their home environment, seamlessly integrate useful data collection into their daily activities and provide patient and caregiver education and reminders.19 20 The first critical step in designing or selecting an acceptable digital health management tool is to ascertain the needs, skills and resources of the target patient audience. We sought to assess access to, utilisation of and facility with technology, as well as the preferred features of a digital health management tool, in patients with an early readmission for decompensated cirrhosis.
Methods
Study design
We conducted a cross-sectional mixed-methods study of patients with an early readmission for decompensated cirrhosis. A semistructured interview obtained both quantitative data through yes/no and multiple-choice questions and qualitative data through open-ended questions. Demographic and health information were obtained from the electronic medical record. Patients were given a study fact sheet and verbal informed consent was obtained. This study was approved by the Partners HealthCare Institutional Review Board.
Study population
We sought to enrol 50 patients with cirrhosis and demonstrated high healthcare utilisation, since this population may benefit most from digital health management tools. We performed daily screening of the inpatient hepatology consult census to identify patients with a clinical diagnosis of cirrhosis requiring admission and readmission within 90 days, for a complication of cirrhosis. Complications were grouped as: (1) hepatic encephalopathy (HE), defined as altered mental status due to chronic liver disease and improved after treatment with lactulose; (2) volume overload, defined as ascites or hepatic hydrothorax, requiring paracentesis, thoracentesis or diuretic adjustment; (3) spontaneous bacterial peritonitis, defined as infection of ascitic fluid and more than 250 polymorphonuclear cells per high power field and/or monomicrobial culture in the ascitic fluid without evidence for an intra-abdominal secondary source such as a perforated viscus; (4) portal hypertensive bleeding, defined as clinically evident bleeding and varices or portal hypertensive gastropathy on endoscopy or (5) hepatorenal syndrome, defined as kidney injury in a patient with cirrhosis and ascites, serum creatinine >1.5 mg/dL, no improvement in serum creatinine after at least 2 days of diuretic withdrawal and volume expansion with albumin, absence of shock, recent nephrotoxic drugs and parenchymal kidney disease. Diagnosis of a complication of cirrhosis was confirmed by the consulting hepatology or the internal medicine teams.
Only patients deemed by their providers to be appropriate for the study were approached. Patients enrolled in hospice were excluded. Other inclusion criteria included age of 18 years or older, English-speaking and capacity to perform informed consent at the time of interview. For patients admitted with HE, a physician investigator evaluated their capacity to consent. If needed, capacity assessment included a Montreal Cognitive Assessment and scores>25 were considered reassuring.
Institutional structure
This study was conducted at a single urban academic liver transplant centre. Patients with cirrhosis are admitted to a general medicine service with hepatology or gastroenterology consultation. The majority of our patients are local and live within an hour’s drive from the hospital, though some patients are transferred from other parts of New England.
Semistructured interview
We created a semistructured interview to target several themes relevant to outpatient cirrhosis management and technology utilisation (online supplementary appendix 1). The themes were generated based on the qualitative research principle of reflexivity: by reflecting on clinical experience, literature review and preconceptions to reduce bias in interviewing and analysis.21 Interview questions explored Smartphone and computer use, personal experience of liver disease, desire to improve health, health data tracking, notifications, caregiver involvement and desired features of a future digital health management tool. Caregiver was defined as someone (a non-professional) with daily contact who aids in managing medical illness. The interview was piloted on the authors and edited in an iterative fashion.
Supplemental material
Study procedures
Once qualified patients provided informed consent, an interview was performed in their hospital room. In-depth semistructured interviews were conducted between June 2018 and January 2019 by one of four investigators (PB, TW, BG, MM) and live transcribed by another investigator. Two investigators were present for every interview. The patient was asked for further clarification when needed. Chart review was performed on all enrolled patients to obtain the following data: demographic factors, aetiology and severity of liver disease and diagnoses at index admission and readmission.
Quantitative data analysis
Responses to individual interview questions were summarised using a mean and SD (continuously measured) or presented as proportions (categorical). Missing data were accounted for by adjusting the denominator.
Qualitative data analysis
Interview transcripts were imported into NVivo 11.0 (QSR International, Melbourne, Australia). Two investigators (PB and TW) from different backgrounds (to minimise personal and disciplinary bias) iteratively read and coded interview transcripts for themes.22 The principle of grounded theory was applied: as themes emerged from the data, specific lines of text were coded into themes.23 The analysts then jointly compared codes, resolved discrepancies and developed a taxonomy of themes. Themes were refined until reaching saturation with a final taxonomy of 16 themes. This final taxonomy was applied to all of the transcripts by the two analysts, with >88% agreement on each theme achieved (0.59 kappa; online supplementary appendices 2 and 3). After this final taxonomy was applied, further subthemes emerged and appear in the tables presented herein.
Supplemental material
Supplemental material
Results
Of 192 cirrhotic inpatients identified from June 2018 to January 2019, 132 were excluded from our study. The majority (99 patients) were excluded due to not being readmitted within 90 days for a cirrhosis-related complication. Out of the 60 patients who met inclusion criteria, 10 declined participation (figure 1). Of the 50 participants, 33 (66%; table 1) were male, mean (SD) age was 57.6 years (11.1), 29 (58%) had alcohol-associated cirrhosis, 14 (28%) non-alcoholic steatohepatitis (NASH) cirrhosis and 4 (8%) viral cirrhosis, with a mean (range) model for end stage liver disease of 22.7 (10–46). On readmission, 20 (40%) patients were diagnosed with HE, 31 (62%) with volume overload and 9 (18%) with bleeding. Patients interviewed had a mix of insurance coverage: Medicare (42%), private insurance (38%), Medicaid (14%) and no insurance (6%).
Patient screening and enrolment.
Clinical characteristics of patients with early readmission for decompensated cirrhosis*
Health attitudes
Patients reported the most bothersome symptoms of their liver disease to be abdominal swelling (72%), fatigue (70%), leg swelling (66%) and confusion (62%). Almost half of the patients (48%) attributed a specific complication as the reason for their admission, while 28% were only able to label their general liver condition as the cause. The remainder did not clearly identify a liver condition as their reason for admission. Twenty-four (48%) patients felt that their readmission could have been prevented (‘yes’ or ‘maybe’). Forty-four (88%) patients reported it is possible to improve their own health (‘yes’ or ‘maybe’).
Comfort with and access to technology
Thirty-eight (76%) patients interviewed own a Smartphone, with an even split between iOS and Android (47% and 45%, respectively). Thirty-one (62%) patients have regular access to a computer with internet (table 2). Of patients who own a Smartphone, 37 (97%) have their Smartphone with them most of the time and 21 (55%) ‘feel like something is missing’ when they do not possess their Smartphone. Smartphone applications are preferred by 66% of patients, compared with 9% who prefer their computer. Eleven (29%) patients have used a health-related Smartphone application, none specific to liver disease. When asked if they would use a Smartphone application for 5 min daily to manage their cirrhosis, 32 (64%) reported yes, 11 (22%) reported no and 7 (14%) reported maybe or not sure.
Patient comfort with and access to technology
A range of comfort with technology was expressed, from significant discomfort to very comfortable and reliant on technology (table 2). One patient reported ‘(I have my Smartphone with me) all the time. That’s my life’; while another stated ‘I’m not tech savvy’. Common themes from the qualitative data include a willingness to learn more about digital tools, a desire to learn about liver disease and communicate with providers with technology and an eagerness to use technology if it could improve health. With regard to learning about digital tools, one patient reported ‘I would have to have someone teach me. Like really teach me. Not just one time. I have to have people come to my level. Then I’ll catch on’.
Caregiver role
Forty-six (92%) patients report having a principal caregiver (table 3). Of the 46 patients with a reported caregiver, 36 of the principal caregivers (78%) participate in illness management and 34 (74%) aid in medication delivery. Thirty-seven (80%) own a Smartphone. When asked if their caregiver would use a Smartphone application to aid in disease management, 71% of patients responded yes, 9% responded no and 21% responded maybe or did not know. Twenty-five (64%) patients would like their caregiver to receive medication reminders (yes or maybe).
Role of the caregiver in managing cirrhosis with digital health management tools
A range of caregiver comfort with technology was expressed, from not comfortable at all to very comfortable (table 3). Common themes from the qualitative data include a reliance on the caregiver, heavy caregiver involvement in medication management, provider communication and management of complications (eg, volume overload and HE). Finally, the caregiver consistently plays a large role in shared learning about the patient’s liver disease. Describing his reliance on his caregiver, one patient stated ‘without her, I’d be dead’. One patient described her family’s management of her encephalopathy: ‘I’m slurring my words and they know. They will get the meds and bring (them) to me’.
Desired features of a digital health management tool
Patients may be interested (answered ‘yes’ or ‘maybe’) in a Smartphone cirrhosis-management application that could: communicate with their physician (39, 85%), send medication notifications to the patient (30, 65%) and the caregiver (25, 64%), transmit diagnostic results and appointment reminders (37, 82%), educate about liver disease (38, 79%) and offer low salt diet recipes (33, 79%) (table 4).
Desired features of a digital health management tool for cirrhosis
Of the 21 patients with an index or readmission with HE, 4 (19%) already play Smartphone or computer games and 14 (67%) would be willing to play a Smartphone game if it could detect cognitive decline. Of patients admitted with volume overload, 85% would be willing to use a Smartphone application to regularly transmit weight data to their doctor; however, 10% of patients felt that would be a privacy violation.
A range of desired features of a digital health management tool were expressed (table 4). Common themes include tools that could provide education, assessment of ongoing ascites and encephalopathy, better communication among providers and a social network with other patients with liver disease. One patient stated: ‘If there were spots (in a Smartphone app) to write in the weights, we would be happy to do that. Because they are always asking us for those things. If I could type that in somewhere and it could go to my doctor, that would be a great thing’. Desired educational topics include effects of each liver medicine, prognosis, life expectancy, common cirrhosis-related procedures and liver transplantation.
Discussion
The medical community has historically questioned the technological skills of patients with decompensated cirrhosis, due to their limited executive function, immobility, age and, in some cases, financial limitations. Instead, this study demonstrated that operating a Smartphone is within their regular scope of activity. We found that the majority of patients with high-risk decompensated cirrhosis in our hospital have a Smartphone, use a Smartphone regularly and would use a Smartphone application for 5 min daily to manage their condition. As such, digital health management tools may be a fertile ground for improving healthcare delivery in this population.
To our knowledge, there is only one prior publication reporting a Smartphone application as an intervention for decompensated cirrhosis.19 In this trial, patients were given an iPhone with the Patient Buddy App installed; hence, the percentage of patients who already owned or used a Smartphone in this study was not reported. The application was used by the patients and caregivers to transmit the following variables to the provider: cognitive function, orientation, risk of falling, medication adherence, daily weights and sodium intake. Only 4 of 40 (10%) patients did not complete application entries during the 30-day trial, largely because they felt it required too much information. The authors reported that eight potential HE-related readmissions were prevented by the application, and most patients and caregivers had a favourable impression of the intervention.
Consistent with this prior work, the majority of patients in our study expressed a willingness to learn about digital health management tools. Despite variable comfort with technology, most patients voiced a desire to learn more about liver disease via an application, communicate with providers via technology and an eagerness to use technology if it could improve their health. The majority of patients with HE would be willing to play a Smartphone game to detect cognitive decline. Likewise, the majority of patients with ascites would be willing to use a Smartphone application to regularly transmit weight data to their provider. Prospective and controlled studies must be performed to comprehensively evaluate the feasibility and efficacy of such technologies.
Caregivers play a central role in disease management for the majority of patients with cirrhosis at our centre. This finding is consistent with the Patient Buddy App trial, which used caregivers as a key part of application implementation.19 In our study, nearly all patients had a caregiver and that caregiver had access to a Smartphone. Most patients suspected that their caregiver would use a Smartphone application to aid in their cirrhosis management. Legitimate concerns about health information privacy have limited providers’ ability to harness the full potential of caregivers, but this study demonstrates that caregivers represent a potentially underused resource in the management of decompensated cirrhosis. Future digital tool interventions should incorporate the use patterns and preferences of cirrhotic caregivers.
It is important to note that nearly one-quarter of patients in this study did not own a Smartphone. Several of those patients reported financial limitations to Smartphone access, not an aversion to the technology and would use a Smartphone application to manage their liver disease if it were given or lent to them at no cost. For centres considering a digital health tool for cirrhosis management, these patients without Smartphone access will require an alternative approach. One avenue, used in the Patient Buddy App trial, is to provide all patients with a Smartphone. In addition, several patients were excluded from study participation for ongoing encephalopathy. Future investigations into digital health management tools for cirrhosis will need to pay special attention to those with encephalopathy.
The results of this study must be interpreted in the context of study design. We can only report the opinions of patients who consented to study enrolment and 10 patients declined study participation. As such, our study may have selection bias for patients more comfortable with technology. Due to study constraints, caregivers did not directly report their access and comfort with technology; rather, this was reported by patients. We also acknowledge that patients’ view of a digital health management tool may vary by setting and change after hospital discharge. Furthermore, the interviews were performed at a single urban centre in New England, potentially limiting the external validity of our findings. Given these potential biases, any reader considering implementation of a digital health management tool should perform a similar survey at their centre to better understand the use patterns and preferences of the targeted population, keeping in mind that tool implementation will likely reveal additional insights about patient preferences.
We believe that the potential value of a digital health management tool lies in its ability to carry on a timely dialogue between the patient and the provider team. The information exchanged certainly has value, like weight or cognitive data in the Patient Buddy App study. Likewise, the cessation of information exchange may also be crucial data and warrant further evaluation and intervention by the provider.
If digital health interventions lead to even a small decrease in readmissions, they will improve the cost of care and quality of life for patients at an anticipated low cost. Among high-cost patients with cirrhosis with an early readmission at our centre, most have Smartphones and would be willing to use a Smartphone to manage their disease. Future technology interventions should be designed keeping in mind the specific use patterns and preferences of this population.
References
Footnotes
Twitter @BloomPringle
Contributors PPB was responsible for the planning, conduct and reporting of the study. PPB is also the guarantor of the study. MM, TJW and BG were responsible for the conduct and reporting of the study. CB was responsible for data analysis. JH and RTC were responsible for the reporting of the study. JMR was responsible for planning, conduct and reporting of the study.
Funding This project was supported by a Partners Healthcare Center of Expertise in Quality and Safety grant and by a Massachusetts General Hospital Department of Medicine Innovation Program Spark Grant, awarded to PPB. PPB is supported by the following grant: NIH T32 DK007191-43. RTC was supported by the following grant: NIH K24 DK078772.
Competing interests None declared.
Patient consent for publication Not required.
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement Data are available on reasonable request.