CASE STUDY : Koha In The Greater Western Area Health Service (N

1

A survey of Koha in Australian special libraries : open source brings new opportunities to the outback.

Introduction

Since 2007, the library management systems landscape has changed markedly, with a variety of open source systems (Breeding 2008b) gaining an increasing share of the market (Breeding 2008c). In Australia, penetration by open source systems has been led by the recent adoption of Koha by an increasing numbers of libraries.

This paper presents a case study of Koha in a rural health library setting, and then presents the results of the first survey of Australian Koha users regarding the viability and practicality of this popular open source system. This survey encompasses both libraries using Koha support companies, and also some which have self-installed systems.

Case Study : Koha in the Greater Western Area Health Service (N.S.W., Australia)

The Greater Western Area Health Service (GWAHS) provides health care to the population of a huge rural area of 444,900 sq. km. or 55.52% of the state of New South Wales, Australia. This is a region more than twice the size of the entire British Isles and larger than the entire state of California. There are 113 health care facilities (52 hospitals and 61 community health centres) spread throughout this vast region.

The library service has five small libraries with which to service a very dispersed clientele. Three libraries are concentrated in the far eastern Bathurst/Orange region (Orange Base Hospital Library; Centre for Rural & Remote Mental Health(CRRMH), Orange; Health Library Service, Bathurst), one in the large regional centre of Dubbo (Dubbo Base Hospital Library) and the fifth (the Far West Health Library) services the remote western area from the mining city of Broken Hill.

Figure 1. Greater Western Area Health Service. Library locations are indicated by the large stars.

By 2006, only one of the 5 libraries in GWAHS (Dubbo) had a web catalogue.There was a general realization that the accessibility of the GWAHS library service would be greatly enhanced if a combined web presence could be established.

There were several problems to be overcome before the goal of a combined web presence could be realized :

· The 5 libraries were operating on 5 different library systems.

· GWAHS was formed in 2005 as a merger of three existing health services. Area service networks have had to be constructed from very differing existing systems. As a result, developing a co-ordinated library service was a slow process.

· GWAHS is an area of large distances and presents logistical difficulties in providing face-to-face service. Using information technology (web catalogues, blogs, email etc ) is a logical approach, but many staff are not on the GWAHS email system. The Intranet is not, therefore, sufficient to reach many library users, and Internet based tools which can be accessed externally are necessary. User feedback indicates that many staff resort to their home Internet, or rural “Internet cafes” rather than use limited work Internet facilities.

· Most GWAHS facilities do not have resident medical staff. Primary health care is provided in concert with a wide range of co-partners, including the Royal Flying Doctor Service, Aboriginal Health Corporations, Divisions of General Practice, and a University Departments of Rural Health. An effective library web presence must be accessible by all of these stakeholders.

· GWAHS has had well documented budgetary problems, and had difficulty finding funds for the project. Similar financial constraints precluded upgrade of any existing systems to web versions.

· The health service server capacity was overloaded and not able to handle new applications.

There were, therefore, certain requirements which were necessary for the project:

1. The combined site had to be Internet based to reach the diverse user clientele.

2. The library system needed to have all basic modules, and be capable of handling data from the five existing systems.

3. Funds were very limited. The system needed to be both cheap and provide value for money.

Why Open source ?

The first approach in the project was to canvass existing vendors to investigate whether any of the existing commercial library systems could be adapted to the purpose of an area-based combined web catalogue. Response was mixed, but two of the existing vendors did submit bids (and one or two other vendors as well). One of these was an innovative solution, whereby the largest three libraries were full members, with the remaining smaller libraries working in OPAC mode only at a minimal cost, updating their holdings by periodic downloads of cataloguing data. However, even with this novel proposal, the commercial bids were still beyond the very limited financial resources available.

The suggestion was made that transferring to an open-source system would resolve the problem of software cost while providing all of the common modules.

Prosentient Systems has a long association with health libraries via the national GRATISNET inter-library loan network. When Prosentient announced that it was now supporting the Koha software, it was seen as an opportunity to achieve the desired result and have competent technical support available. Other open source systems such as Evergreen were investigated, but the availability of local technical support for Koha was a very important factor.

Hosting of the GWAHS Libraries Network

In the case of GWAHS, there were cogent reasons for an externally hosted system :

· Koha runs best on the Linux platform. This is not supported by the NSW Health Windows-based environment.

· The health service server capacity was inadequate to mount new applications.

· Many library users were external, and not reached by the health service intranet.

· Health service IT support staff had little familiarity with open source products.

· Customizations and development of the software were difficult in the health service environment.

· Reliable support companies were available.

· One annual support fee (lower than for the previous vendor) which simplifies the business case considerably, as internal costs do not need to be calculated.

Installation

In April 2008, the conversion to Koha2 was commenced.

The GWAHS task of migrating data from 5 systems into a sixth had very few problems. One of the systems left duplicate MARC records for multiple copies, but otherwise data transfer was clean. Data conversion costs were very economical. Patron data likewise transferred without problems.

The willingness of the support company to conduct training at one of the rural sites was very helpful, allowing system parameters, network rules and similar matters to be dealt with in one sitting. The GWAHS Libraries Network installation went live in August 2008.

Links to the GWAHS Libraries Blog, PubMed and the Health Service website formed part of the customization.

Figure 2 : GWAHS Library Service main page

The system was upgraded to Koha3 in early 2009.

Koha : challenges and opportunities for GWAHS.

The switch from in-house systems to a web-based open source library system, is a great advance for GWAHS, but is not entirely trouble free. The following points are of note.

· Web-based systems need good Internet access to provide optimum benefits. Many parts of GWAHS are very isolated, and Internet response times vary enormously. Opac times in Koha seem to be largely acceptable, but Cataloguing and Circulation modules are often difficult to use if the health service network is running slowly. The recent development of a PC version of the Circulation module (which is regularly downloaded to the main database) may greatly assist this.

· Broadband coverage in rural areas of Australia remains patchy. Slow speed and narrow bandwidth remain problems for GWAHS users. However this problem is insignificant compared to the former problem of studying in a remote location without direct desktop access to library services. The Federal Government Australian Broadband Policy initiative will soon reach some areas of GWAHS and hopefully lead to upgraded Internet access.

· Professional development for GWAHS library staff is necessary to implement change and to keep up-to-date with rapidly developing web technology, Conference attendance and meetings are often difficult and costly to arrange. Teleconferencing and videoconferencing are used, but are a poor substitute.

The outcome for GWAHS

For GWAHS, Koha is an example of open source technology which can be readily and successfully implemented even in an environment where information technology is basic and funds are limited. Change can be successfully achieved through the networking of resources and a willingness to confront the challenges of serving clients spread over a vast area. Balnaves and Keast (2009a) discuss the process in more detail.

Spread of the Koha concept to Australian special libraries.

The Greater Western AHS Library Network (July 2008 : 5 libraries) was the first network of Australian health libraries to move to Koha. The experiment aroused considerable interest in the health library community. GWAHS was soon followed by the South Eastern Sydney Illawarra Area Health Service Libraries Network (January 2009 : 11 libraries) and the southern part of the Northern Sydney Central Coast Area Health Service Libraries Network ( August 2009 : 5 libraries). The potential exists with open source for small library networks to put their libraries on the Web at minimal cost. Several isolated health libraries also made the move.

Adoption of Koha by other Australian special libraries has also spread rapidly since 2008 (although isolated in-house installations date as far back as 2004). The largest by far of these is the recent addition of the Australian Bureau of Statistics National Library Services (March 2010).

There are now upward of 45 Australian libraries using Koha. The added functionality of the Koha3 release has made the system much more viable and attractive.

Survey of Australian Koha users

In order to ascertain the reasons why libraries made the shift to Koha, and what they actually thought of the performance of the system, the author conducted an Internet-based survey of Australian Koha libraries (Keast 2010) in February 2010, an expanded version of an earlier survey conducted in June 2009 (when several of the respondents had only recently gone live). The survey was forwarded to all Australian Koha libraries which could be traced. The response rate was 21/45 or 46.67%. The respondents were health & medical libraries (71%) and other specials (29%). The respondents included clients of both major support companies, and 7 in-house installations. There was little difference in responses between any of these groups. The results of this survey are discussed below.

Why change library systems ? :

For a few libraries, installing Koha provided a library system, something which they could not afford before, However most libraries surveyed did have a pre-existing system (85.7%). 9/21 formerly had a Web catalogue.

Two main reasons were given in the survey for changing to open source systems :

1. Dissatisfaction with previous systems, especially the lack of flexibility in achieving customizations.

“It didn't do what we needed it to do; in particular, reporting was difficult. There was no flexibility in the system, and lots of new "fixes" were obviously queued forever.”

“Software very old, inflexible. Poor support, expensive maintenance fees.”

There was a large amount of dissatisfaction with existing conventional systems, and some library experiences were unpleasant and definitely not user-friendly. Some examples :

· The U.S. parent company decreed that all users should purchase all upgrades. A library with a minimal budget was left with a bill for $A6000 for an upgrade it did not need and could not afford.

· A very adequate Australian based small library system, suffered a loss of technical support after its founder died. The company was subsequently sold to a new owner. The new company merged with a U.S. based company, who announced that they would no longer support the product. Users were offered an upgrade to another new product (which turned out to be heavily school-focused and unsuitable for medical or technical collections).

· “The previous vendor quoted $10,000 to extract our data from its system which was too expensive for a small organisation and also seemed somewhat excessive for only 1,600 titles.”

2. Budgetary difficulties.

The global financial crisis has been felt by many small libraries, and many health services as well. Upgrading existing systems, graduating to web-based catalogues, and paying for customized developments are simply beyond the resources of many small libraries. Koha is seen as a low-cost alternative, with lower maintenance costs, and considerably lower start-up costs. Jeffrey (2010) in an interesting internal paper written from a commercial system vendor viewpoint, points out certain trade-offs that making the open source decision on cost grounds involves.

o Dedicated support from a support company, or considerable in-house expertise is still needed, and these functions cost in both staff time and expertise.

o Support will generally come from an IT professional rather than a product specialist

o Regular updates are not guaranteed and are dependent on the whims of the “community”

o There is a lack of quality assurance ( a good example was a garbled and incomplete set of manuals for Koha2, largely addressed in Koha3)

o With many developers involved, there is an overall lack of direction for open source products.

o While customizations may be easier with open source, this means that there are inconsistencies in the product between installations, source code may become compromised, and there is little clear direction as to which customizations will (or should) find their way into the overall product.

Most of these points have some validity and provide food for thought. However, for many of the libraries surveyed two questions only were paramount (a) Does this system satisfactorily deliver basic functions of a library management system ? (b) Does it do this at a lower and affordable cost ? This clearly mirrors the situation in GWAHS, where the critical question was : “Could a combined web presence be achieved with an overall reduction in cost ?”

Why Koha ?

While local support and cost savings were strong motivations to move to Koha rather than other open source systems, most libraries conducted a thorough investigation into Koha’s range of functions and flexibility, and that of its open source competitors. Balnaves (2008) gives a succinct analysis of criteria for evaluating open-source systems, and applies these criteria to seven of the most well-known open source systems. Bissels (2008) gives a detailed account of the demanding evaluation process for the UK-based CAMLIS network [which adopted Koha3].

Hosting and support :

Breeding , writing in 2008 (Breeding 2009) states “ In the current environment, almost all libraries implementing open source ILS products do so with paid support contracts “. In 2008, this was certainly also true of the Australian scene (although 1 in-house tertiary installation dates to 2004). However, since then, increased functionality and better manuals in the Koha3 release have seen a number of in-house installations in Australia, but these are mainly in tertiary organizations or government organizations with good IT departments.