Elsevier

Information Systems

Volume 71, November 2017, Pages 111-122
Information Systems

Power to the patients: The HealthNetsocial network

https://doi.org/10.1016/j.is.2017.07.005Get rights and content

Highlights

  • We defined a social network that helps patients to make more informed choices related to their health status.

  • We implemented a component that supports patients to identify the most relevant medical areas for their health problems.

  • We implemented a recommender system that facilitates patients to find a doctor, a health facility, or a hospital for treating their health problems.

  • We performed experiments whose results show the effectiveness of the approaches.

Abstract

HealthNet (HN) is a social network that brings together patients with similar health conditions. HN helps users in finding a solution to their health problems by suggesting doctors and health facilities that best fit the patient profile. Indeed, the core component of HN is a recommender system that suggests patients similar to the target user and supports the choice of the doctor and the hospital for a specific condition. The recommendation algorithm first computes similarities among patients, and then generates a ranked list of doctors and hospitals for a given patient profile by exploiting health data shared by the community. The HN typical user can find the most similar patients, can look how they treated their diseases, and can receive suggestions for solving her condition. In order to facilitate the interaction with the system and improve the recommendation step, the patient can express her health status by a natural-language sentence. The system analyzes the sentence and identifies the most relevant medical area (e.g., orthopedics, neurology, allergology, etc.) for that specific case, and uses this information for the recommendation task. Currently HN is in alpha version and only for Italian users, but in the future we want to extend the platform to other languages. We carried out both an in-vitro experimental evaluation to assess the effectiveness of the module for analyzing natural language descriptions provided by users as well as the recommender system to suggest the right doctors for a specific health problem, and an in-vivo evaluation performed by real doctors. Results are really encouraging.

Introduction

The health status is one of the aspects which mainly affects the quality of a person’s life [26]. When needed, finding a physician able to properly diagnose and treat a medical condition is a very tricky phase, and family or friends usually represent the very first source of information for seeking the right doctor or health facility. E-health, defined as the healthcare practice supported by the electronic process and communication [13], is revolutionizing this scenario and the landscape of clinical practice. A recent survey demonstrated that 72% of U.S. Internet users looked online for health information within the past years.1 Similarly, in Italy, 84% of young people between 18 and 35 uses the Web as a source of health information.2 This new phenomenon can be viewed as the evolution of the process that a patient plays out in order to find a solution to her conditions, i.e. to ask the family or friends who had the same problem how they treated it. Indeed, in the same aforementioned survey, 60% of U.S. adults got information or support from friends and family when they have a health problem and 24% of adults got information or support from others who have the same health condition.

A further proof of this ongoing (r)evolution was the theme of the 19th International Society for Quality of Life Research (ISOQOL) Annual Conference: ”The Journey of Quality of life research: A Path Towards Personalized Medicine”. One of the four plenary panels was ”Innovations in e-health” [51]. The general thread shared by all talks was the exploitation of e-health tools for empowering individuals to be actively engaged in the management of their health. Indeed, the sharing of information generates a more informed and empowered patient by reconfiguring the patient/care team relationship towards a patient-centered medicine. One of the most relevant initiatives in that direction is the U.S. social network PatientsLikeMe (PLM)3 [7], which enables patients to share and compare different diagnoses and treatments with people having the same conditions anywhere in the world [2]. PatientsLikeMe counts 300,000 patients sharing 2300 different conditions. Furthermore, there are a lot of forums, blogs, and more generally web sites which deal with health problems, but the information is often confused, difficult to understand and can lead to easy, often wrong self-diagnosis [25]. Indeed, the enormous diffusion of sites that disseminate medical information on the Web has contributed to the emergence of a phenomenon called cyberchondria, i.e. the unfounded perception of concerns about common, likely innocuous symptoms that can escalate into the review of content on serious, rare conditions that are linked to the common symptoms [50]. In this context we think that the role of the Web in the health domain should encourage the connection between the patients and the doctors that are the only persons able to provide a correct diagnosis.

In this paper we present HealthNet (HN), a social network whose main goal is to help users to find a solution to their health conditions by facilitating the connection between patients and doctors. The choice of a doctor or a health facility is a typical problem of information asymmetry where the patient has too little information to make an informed choice and thus needs to be somehow supported. Accordingly, HealthNet helps patients to share knowledge, find other similar patients, in order to look at their experiences.

HealthNet implements: (1) a self-tracking system that allows users to trace their physical and biological parameters, (2) a Natural Language Processing module able to understand the user health status expressed through natural language sentences, (3) a recommender system that allows to discover the similarity between patients and to exploit the data coming from the community of patients for suggesting doctors, health facilities, or hospitals that best fit the patient profile. This should help to deter the self-diagnoses.

The main contributions of the paper are the following:

  • we defined a social network which helps patients to make more informed choices related to their health status;

  • we implemented a component that supports patients to identify the most relevant medical areas for their health problems;

  • we implemented a recommender system that facilitates patients to find a doctor, a health facility, or a hospital for treating their health problems.

The rest of this paper is organized as follows. Section 2 provides a comparison between HealthNet and other relevant researches proposed in the literature. Section 3 describes the platform and its general architecture, while Section 4 provides some typical use cases. The experimental evaluation is described in Section 5, and Section 6 draws conclusions and future work.

Section snippets

Related work

E-health – also known in the early stage of the research as health informatics or consumer health informatics – is a wide research area embracing personalized medicine, health social networks, health recommender systems, digital prevention and interventions, smart health. In [6], Brennan and Safran already argued that consumer health informatics innovations provide information about patient health concerns, assist users to find others who share their concerns, afford platforms to promulgate

The HealthNet social network

HealthNet is a social network where users are patients. In addition to the traditional services offered by social networks, HN provides a set of functionalities, such as a self-tracking system, a PHR manager, and a recommender system. HN provides a user-friendly interface that allows an easy access to each functionality. The first step of the interaction with the system is the user registration. After that, the patient can start to record her medical data. In order to preserve the privacy, the

Typical use cases

In this section we define some typical use cases of HealthNet. We identify three main scenarios: the recommendation of a hospital, the recommendation of a doctor, and the sharing of an experience.

Experimental evaluation

The goal of the experimental evaluation is twofold:

  • to assess the effectiveness of the process of automatically identifying the medical area for health problems described using natural language;

  • to assess the effectiveness of the system to recommend the right doctors for a specific health problem;

  • to assess the accuracy of the recommender system through a validation performed by domain experts.

More specifically, in experiment 1 we test the accuracy of different classification algorithms and

Conclusions and future work

In this paper, we presented HealthNet, a social network where patients can share their experiences with the community, such as consulted doctors, hospitalizations, health facilities. The goal of the platform is to use this information for suggesting doctors and health facilities which best match a given patient profile. Accordingly, the core component of HealthNet is a recommender system able to find similar patients based on the description of their symptoms, conditions and treatments, in

References (55)

  • R. Burke

    Hybrid recommender systems: survey and experiments

    User Model. User-Adapt. Interact.

    (2002)
  • E. Cardillo et al.

    A methodology for knowledge acquisition in consumer-oriented healthcare

    Knowledge Discovery, Knowledge Engineering and Knowledge Management

    (2009)
  • A. Cawsey et al.

    Adaptive information for consumers of healthcare

    The Adaptive Web

    (2007)
  • L.A. Celi et al.

    Disrupting electronic health records systems: The next generation

    JMIR Med. Inform.

    (2015)
  • D. Cox

    The regression analysis of binary sequences (with discussion)

    J. Roy. Stat. Soc. B

    (1958)
  • V. Della Mea

    What is e-health (2): The death of telemedicine?

    JMIR

    (2001)
  • L. Duan et al.

    Healthcare information systems: data mining methods in the creation of a clinical recommender system

    Enterp. Inf. Syst.

    (2011)
  • C. Ehrentraut et al.

    Detecting healthcare-associated infections in electronic health records: evaluation of machine learning and preprocessing techniques

    Proceedings of the Sixth International Symposium on Semantic Mining in Biomedicine (SMBM 2014)

    (2014)
  • R. Eriksson et al.

    Dictionary construction and identification of possible adverse drug events in danish clinical narrative text

    J. Am. Med. Inf. Assoc.

    (2013)
  • R.G. Farrell et al.

    Intrapersonal retrospective recommendation: lifestyle change recommendations using stable patterns of personal behavior

    Proceedings of the First International Workshop on Recommendation Technologies for Lifestyle Change (LIFESTYLE 2012), Dublin, Ireland

    (2012)
  • L. Fernandez-Luque et al.

    Challenges and opportunities of using recommender systems for personalized health education.

    MIE

    (2009)
  • R. Freeman et al.

    Advances in electronic surveillance for healthcare-associated infections in the 21st century: a systematic review

    J. Hosp. Infect.

    (2013)
  • M. de Gemmis et al.

    Semantics-aware content-based recommender systems

    Recommender Systems Handbook

    (2015)
  • M.d. Gemmis, L. Iaquinta, P. Lops, C. Musto, F. Narducci, G. Semeraro, Learning Preference Models in Recommender...
  • T.D. Gunter et al.

    The emergence of national electronic health record architectures in the united states and australia: models, costs, and questions

    J. Med. Internet Res.

    (2005)
  • A. Hartzler et al.

    Managing the personal side of health: how patient expertise differs from the expertise of clinicians

    J. Med. Internet Res.

    (2011)
  • B.W. Hesse et al.

    Trust and sources of health information.

    Arch. Intern. Med.

    (2005)
  • Cited by (11)

    • Multi-label clinical document classification: Impact of label-density

      2019, Expert Systems with Applications
      Citation Excerpt :

      This paper faces the challenge of classifying a particular type of documents, Electronic Health Records (EHRs), that is, patient records created on a large scale in the daily routine of the hospitals. Clinical text mining, a sub-field of NLP, is applied to discern knowledge from clinical texts (Dwivedi, 2016; Narducci, Lops, & Semeraro, 2017). Notably, we deal with the classification of EHRs according to the International Classification of Diseases (ICD).

    • Healthcare Providers Recommender System Based on Collaborative Filtering Techniques

      2022, Machine Learning and Deep Learning in Medical Data Analytics and Healthcare Applications
    • Health recommender systems: Systematic review

      2021, Journal of Medical Internet Research
    View all citing articles on Scopus
    View full text