Socio-Economic Aspects of Healthcare and Mobile Search

Authors: Antonio Foncubierta Rodríguez (HES-SO, CH) and Henning Müller (HES-SO, CH)

Contributor: Shara Monteleone (JRC-IPTS, ES)

Table of content

1. Introduction

2. Methods

3. Search Market Main Vendors

4. Available Complete Solutions

5. Conclusions

6. References

 Introduction 

This document is created as part of the Chorus+ project that mainly deals with multimedia retrieval systems. Two focus areas of the project are aspects of mobile search and enterprise search. For mobile search a document has been made available [1], whereas for enterprise search the document is current being finalized.

Health information search can be regarded in both mobile and enterprise aspects and thus this document is an annex to the two cited domains and document son enterprise and mobile search.

Health information search can in many cases be regarded as an application of enterprise search, since it shares some of the limitations and requirements of enterprise search as opposed to generalist search in often openly accessible data.

In healthcare environments as well as in industrial environments internal policies limit the availability of information based on the user role or permissions. For instance, depending on national regulatory frameworks, electronic health records can only be accessed by a restrained number of persons, often persons directly involved in treating a specific patient, similarly to how internal content (such as the intranet but also mail traffic or created documents or automatically created content such as web log files) in companies can only be accessed by persons with defined roles in the company.

Mobile devices have rapidly spread among general public but only in recent years their use has moved from basic voice/text point-to-point communications to more complex tasks. Mobile information search has risen as one of the applications that attract both the interest of users and service providers. Context-aware search applications have empowered users by providing information relevant not only to the query but also to the context where the search is performed. A detailed description of mobile search can be found in [1]. Mobile medical search is presented in [2].

In this report, we focus on the health domain, where the term mobile health is understood as the use of mobile devices to query health related information, whether it comes from the electronic health record, web sites or other sources. More specifically this report deals with health search applications meeting also enterprise search requirements, such as role-based access, so also on internal and potentially protected documents.

Methods

For the elaboration of this report, a selection of the main vendors in the search market was needed. This selection was carried out based on various external analyst reports on the topic [3] [4] [5]. Being a market that evolves rapidly, some vendors might not be included in this report as an exhaustive analysis was not the main objective. For each of the vendors present in these reports, their products or results were studied in relation to the three domains with which this document deals: enterprise search (and how this is integrated), health information and mobile search (with details on the types of interfaces).

Figure 1 Classification of enterprise search vendors, concentrating on the application areas and not on specific platforms of mobile devices.

Search Market Main Vendors

The search market is a very active area of business and research, where many players interact and offer their services to companies and individuals. Figure 1 shows a broad classification of some of the most important providers/vendors in the field, focusing on the interaction of two or more markets, leaving out single-market solutions. The classification in Figure 1 does not distinguish between solutions that aim at several markets simultaneously and vendors that have specific solutions for each market. This distinction will be made in section 4, where the main solutions are discussed.

In Table 1, a description of the main products of the selected vendors is shown. For each of the vendors, search products and healthcare related products were investigated.

Table 1: Summary of the main vendors in enterprise search and their products.

Available Complete Solutions

As can be seen in section 3, not all vendors provide solutions for the three markets that this text targets. Moreover, even the vendors competing in the three fields together do not provide a single solution for all three areas. In this section, the solutions that can be considered complete, meaning that they address all three areas with one single product, are discussed. This analysis gives a general view of what the market niches are that are yet to be exploited.

Autonomy Auminence

Being one of the leaders in the Enterprise Search market, Autonomy provides a solution that empowers clinicians with knowledge for clinical decision support. Focused on diagnosis, Auminence provides a list of possible diagnosis, recommended tests as well as integration with electronic health records.

Figure 2 Screenshot of the Autonomy Auminence application.

Strengths:

Very complete solution.

Integrated tool with an appealing interface.

Cautions:

Free-text query entry is limited to symptoms.

Relies mostly on structured information: symptoms, age, sex, etc.

Lacks a multimedia/multimodal approach.

Epocrates

Epocrates has a long history providing access to medical content from mobile devices: it reached popularity among healthcare professionals in the late 1990s with a Palm application. Aimed at clinicians since the acquisition of Modality, it contains now more than 65 applications for improving the clinicians’ decisions at the point of care.

Figure 3 Screenshot of the Epocrates Android application.

Strengths:

Specific content search engine: drugs, diseases and medical calculator among others.

Well-known among USA healthcare professionals.

Portfolio of iPhone applications from medical vocabulary translation to imaging atlases, also Android applications.

Cautions:

Lack of a content-based query approach.

Lack of an integrated search with several result types. Every type of content is queried separately.

Khresmoi

In one year since the project kickoff, the Khresmoi project has been approaching a complete search solution for clinicians and the general public based on general health data, so excluding clinical records. By identifying various use cases, Khresmoi tries to empower the health knowledge of the general public s well as the clinicians’ decision making with a multilingual and multimodal search engine.

Strengths:

Multilingual and multimodal approach.

Content-based retrieval for visual search.

Restricted data access for clinicians and radiologists.

Public data access for the general public.

Mobile prototype.

Cautions:

Early stage prototypes.

No search in clinical data and external sources combined.

Business model still to be defined, due to the preliminary stage of the project.

Microsoft

With two information-management tools, Microsoft approaches the triple problem of mobile, health-related, enterprise search. However, it is still an incomplete approach since the mobility it provides relies on certified devices for Microsoft HealthVault and its main feature is synchronization with an in-the-cloud personal health record storage. Information access capabilities other than indexing/browsing are not defined in commercial information about HealthVault. It includes a “Web Health” search engine.

Figure 4 Screenshot of Microsoft Health Vault.

Strengths:

Supported by a well-known company, can push its adoption.

Certification system for devices.

Integration with third-party online tools to share and track health information.

Cautions:

Only present in the US and its territories.

Limited mobility.

Aimed at the general public, privacy issues can be problematic.

Vivísimo

Although Vivísimo does not have a specific healthcare solution, various health providers are using Vivísimo’s Velocity product actively. Velocity is an information optimization platform that uses several machine learning tools as well as semantics and knowledge bases to provide relevant results for enterprise search.

Strengths:

Interesting result optimization by using refinement, automatically learned topics and other semantic techniques.

Mobile extension for the Velocity platform.

Cautions:

Lack of multimodal/multimedia approach.

Not a specific product, but tailored tuning depending on the customer: less visibility.

A comparison of the features provided by each solution is shown in the following table:

Table 2 Summary of features of systems that address several of the search dimensions.

Conclusions

Although an important number of vendors participate in the enterprise search, health search and mobile search markets, in the elaboration of this report it was found that only a limited number of them can target all three fields with one single solution.

The first conclusion of this report is that among the vendors who provide a closed solution for the challenge, the search features are often naïve: relying on structured information and keyword-based queries can be a valid approach for clinicians, but not necessarily for the general public, which lacks prior knowledge on health. This problem specially affects Microsoft’s HealthVault and Khresmoi solutions, since both target the general public. Microsoft partly solves this problem with the integration of input sources: pharmacies, clinicians and devices can put information into the patient’s vault. Khresmoi plans to solve this problem through the general public use case.

A second conclusion of this document is that multimodality is often neglected as one of the sources of health information. Only Khresmoi targets the search problem with visual information as well as textual queries.

Finally, an important remark is that mobility is still an aspect to be explored, since most of the efforts in this field are limited to providing a mobile interface to access the system, without exploiting the context-aware capabilities of handheld devices.

References

x