Summary table
What was already known on the topic?
- •
Health workers have
Information systems (IS) have the potential to address numerous problems in healthcare delivery, patient safety, and clinical practice. This may be especially so if IS can be successfully implemented at the point of care [11], [17]. Unfortunately, health workers have numerous concerns about IS usage and its implications for their work [5], [6], [7], [37], [50], [59], [58]. Problems with health workers' acceptance and satisfaction are now regarded among the most significant barriers to the diffusion of IS within health settings [15], [34], [53]. Tracking the needs of health workers and understanding the attributes of an IS that are most important to their satisfaction and productivity have thus become important components within health IS evaluation and research [4].
An IS user's evaluation of system attributes can influence their affective satisfaction and in turn their usage behaviours [19]. More satisfied users have been associated with deeper levels of engagement with a system's functionality [3], [31], [35], which is important to achieving higher-order benefits from IS implementations [20]. In the hospital context, nurses comprise the largest group of workers and, as generators and users of health information, need to interact frequently with hospital IS systems [61]. Therefore, improving nurses' satisfaction with hospital IS (HIS) and ensuring that HIS embed attributes necessary for nurse productivity are critical issues for hospital administrators and HIS providers. However, there are few studies in the hospital context that examine how multiple IS system attributes correlate with both user satisfaction and productivity outcomes for nurses (see [12], [29]). Past studies have also not adequately examined the relative importance of different HIS attributes to nurse-user outcomes in a manner that distinguishes between those attributes for which underperformance is acceptable to nurse users, those attributes where high performance must be maintained, and those attributes that must be prioritized for intervention. One recent attempt to prioritise HIS attributes from the nursing perspective used a fuzzy analytic hierarchy approach [36]. However, this approach is based on a very small sample size and does not identify priorities with reference to a criterion variable. Without understanding the system attributes most important to users' satisfaction and productivity, efforts to improve HIS performance cannot be effectively prioritized. Consequently, there is risk that the intended impacts of HIS on nursing outcomes will be compromised.
The purpose of this study is therefore to identify priorities for managerial intervention by determining which attributes of an integrated hospital information system (HIS) are most important to the satisfaction and productivity of nurses who use the system in day-to-day clinical practice. Our study is undertaken within an importance-performance analysis (IPA) framework [42]. We utilize both quantitative and qualitative data collected from a public hospital in South Africa to understand nurses' evaluations of the performance of the hospital information system along the identified attributes, and the relative importance of these attributes to nurses' satisfaction with the system and its impacts on their productivity. Our results identify those system attributes requiring the most immediate attention, which can help health care administrators to direct scarce resources in a manner that reduces the risks of system failure, user frustrations and resentment.
The attribute-level performance of an IS has long been recognized as fundamental to user satisfaction and IS success outcomes (e.g. [9], [19], [33], [62]. The DeLone and McLean model of IS success provides a particularly useful organizing framework for understanding system attributes relevant to users. According to this model, attributes along dimensions of technical system quality, stored data quality, information output quality, and support service quality are important to a user's evaluation
The empirical setting for our study is a public hospital in South Africa. The hospital has regional hospital status and had approximately 570 beds at the time of data collection with a staff complement of about 2200. The hospital services include general surgery, orthopaedics, ENT, ophthalmology, general medicine, radiotherapy and oncology, coronary care, high care, obstetrics/maternity gynaecology, paediatrics, as well as a number of support services including inter-alia radiology, laboratory,
Ninety percent of the respondents were female, which is generally reflective of the nursing population in South Africa. Over 50% of the nurses have more than 10 years of experience in nursing. About 40% have spent more than 10 years at the hospital under study, with 90% having spent at least four or more years at the hospital. The nurses surveyed cover a large area of hospital activity. Of the responses, 34.4% were from general medicine wards, 11.7% from surgical wards, 8.4% from intensive and
Health workers have numerous concerns about IS usage. Addressing these concerns requires understanding of the system attributes most important to their satisfaction and productivity. Our study examined nurses' evaluations of the performance of an integrated hospital information system along a set of attributes representing system quality, data quality, information quality, and service quality. We also considered the relative importance of these attributes to nurses' satisfaction with using the
No conflicts to declare.
Authors contributed equally to the conceptualization of the study. Cohen wrote the paper and was predominantly responsible for carrying out the quantitative analysis. Coleman wrote the paper and was predominantly responsible for carrying out the qualitative analysis. Kangethe was responsible for data collection and this work formed part of his Master's research studies. Cohen and Coleman also finalized the paper for submission. Summary table What was already known on the topic? Health workers have
The study was supported in part by funding provided to the first author from the National Research Foundation, South Africa.
This study adopted a pragmatic approach using a sequential mixed-methods design that comprised a qualitative study to inform questions for a quantitative study. Studies on health sciences have previously used mixed research methods to achieve an accurate and comprehensive interpretation from empirical research (Cohen et al., 2016; Campos et al., 2017). This exploration is conducted in three phases.
In addition, 17% of them had worked in the library for less than 5 years, 19% for 6–10 years, 12% for 11–15 years, 22% for 16–20 years, and 30% for more than 20 years. The scales used in this study consisted of three sections, which were adapted from previous studies, including the scales developed by Cohen et al. (2016) and Wang and Wang (2009). The first section involves university librarians' demographic data.
Allocation of further budget and considerable effort is fundamental to improving these attributes. The application of Importance – Performance Analysis (IPA) for deriving managerial actions has been used in various fields including tourism (Boley et al., 2017; Lai and Hitchcock, 2016; Pan, 2015; Randall and Rollins, 2009; Ziegler et al., 2012), healthcare (Abalo et al., 2007; Cohen et al., 2016; Hawes and Rao, 1985; Mohebifar et al., 2016), education (Alberty and Mihalik, 1989; Nale et al., 2000; O’Neill and Palmer, 2004; Phadermrod et al., 2019), public administration (Hua and Chen, 2019; Keith and Boley, 2019), banking (Joseph et al., 2005; Matzler et al., 2003), e-business, technology and IT (Chen and Ann, 2016; Magal and Levenburg, 2005; Skok et al., 2001), food services (Tontini and Silveira, 2007), the automotive industry (Matzler et al., 2004), the building industry (Lee, 2019) and sport centres (Rial et al., 2008). Validity and reliability issues associated with IPA have also been reviewed in previous researches (Azzopardi and Nash, 2013; Bacon, 2003; Griffin and Hauser, 1993; Gustafsson and Johnson, 2004; Lai and Hitchcock, 2015; Oh, 2001; Sever, 2015).