Skip to main content
NCBI home page
Alcohol and Alcoholism (Oxford, Oxfordshire) logoLink to Alcohol and Alcoholism (Oxford, Oxfordshire)
. 2024 Nov 13;59(6):agae075. doi: 10.1093/alcalc/agae075

Investigating the service utilization and pathways of patients with alcohol use disorders

Jingxiang Zhang 1,2,2, Siyu Qian 3, Guoxin Su 4, Chao Deng 5, David Reid 6, Barbara Sinclair 7, Ping Yu 8,✉
PMCID: PMC11558449  PMID: 39535132

Abstract

Background

Process mining was widely used in healthcare settings to analyse patient pathways, providing insights about patient care and resource allocation. This study aimed to apply process mining to explore service utilization and pathways of patients with alcohol use disorders in receiving the community-based Drug and Alcohol Service to establish a baseline for evaluating improvement initiatives.

Methods

Patients’ encounters with the Drug and Alcohol Service and their direct care activities (i.e. Intake, Assessment, and treatment service activities) that occurred in these encounters were analysed to depict the actual pathways. Patterns were uncovered from these actual pathways using a process mining tool Disco. Patients with alcohol use disorders were divided into two groups—patients who use polysubstance and those who do not—to compare their service utilization and pathway patterns.

Results

Five hundred seventy-one patients with alcohol use disorders incurred 1447 encounters, comprising 13,974 activities. The top three actual pathways with at least three activities were: (i) pathway ‘Intake-> Withdrawal Management-> Withdrawal Management’ (170 (11.7%) encounters and 130 (22.8%) patients); (ii) pathway ‘Intake-> Adult Counselling-> Adult Counselling’ (161 (11.1%) encounters and 126 (22.1%) patients); and (iii) pathway ‘Intake-> Assessment-> Withdrawal Management’ (155 (10.7%) encounters and 129 (22.6%) patients). However, 313 (21.6%) encounters did not proceed beyond the Intake stage. When the patients started their pathways from Intake, their immediate next activities were frequently Withdrawal Management, Assessment or Adult Counselling. Compared to those who do not use polysubstance, patients who use polysubstance had a higher proportion of Intake (19.3% vs 11.8%, P < .001), Involuntary Treatment (4.0% vs 1.0%, P < .001) and Magistrate Early Referral into Treatment (6.4% vs .3%, P < .001), but a lower proportion of Withdrawal Management (31.2% vs 39.0%, P < .001) and Adult Counselling (27.8% vs 37.6%, P < .001).

Conclusion

This study contributes to an in-depth understanding of service utilization and pathways of patients with alcohol use disorders within the community-based Drug and Alcohol Service. Future studies should investigate their reasons for dropping out after Intake.

Keywords: alcohol use disorders, patient pathway, process mining, drug and alcohol service

Introduction

A patient pathway can be defined as a series of clinically relevant events that occur during a patient's healthcare journey (Kwok et al. 2021). Patient pathway analysis focuses on the flow of patients throughout the healthcare system and the barriers at each stage, including patient identification, diagnostic testing, treatment decisions, and prescribing (Kwok et al. 2022).

Digital tools, such as electronic health records, are commonly introduced in healthcare facilities, including community-based Drug and Alcohol Service, to standardize and centralize information management in Australian public health systems (Hillestad et al. 2005; Munyisia et al. 2017). However, the high volume and complexity of electronic health record data hinder their use in secondary analysis for understanding healthcare operational management (Sun and Reddy 2013). Moreover, the dataset does not automatically provide easily understandable visual demonstrations of patient pathways, e.g. process, time, and cohort. Therefore, the researchers resort to process mining techniques to uncover insights from patient pathways within a healthcare system (Antonelli and Bruno 2015).

Process mining is the technique of mining knowledge from existing event logs in data sets, such as electronic health records, to discover, monitor, and improve actual work processes (Van Der Aalst 2016). It can provide detailed, valuable clinical insights about health service performance, from information about the sequence, duration, and resources used in real-world patient care activities (Rojas et al. 2016). These insights can support decision-makers in healthcare services to further analyse and improve the task arrangement and resource allocation to improve clinical outcomes (Arias et al. 2020).

Process mining has been applied to a variety of clinical settings across diverse healthcare domains utilizing different techniques and datasets (Rojas et al. 2016). Notable examples include discovery and analysis of patient pathways (Gabrielle dos Santos et al. 2021), comparison of discovered patient pathways with real-world data or other discovered patient pathways (Kim et al. 2013; Partington et al. 2015; Rovani et al. 2015), assessing the extent to which clinical guidelines are followed (Caron et al. 2014; Xu et al. 2020), repairing or improving existing patient pathways (Rovani et al. 2015; Canjels et al. 2021), and prediction of patient activity based on processes performed (Metsker et al. 2019; Cho et al. 2020). The proper and effective implementation of healthcare processes can significantly improve the quality of hospital services and patient experience (Rojas et al. 2016).

The Drug and Alcohol Service offers multiple benefits for public hospitals in Australia. These include reducing emergency department (ED) frequent ‘fliers’, waiting time and length of stay, preventing ED performance deterioration, and reducing hospital admission rates and re-admissions from patients with substance use disorders, including alcohol use disorders (Butler et al. 2016; Reeve et al. 2016; Lintzeris et al. 2020).

The Drug and Alcohol Service at the Illawarra Shoalhaven Local Health District (n.d.) (ISLHD), New South Wales (NSW), Australia provides a diverse range of community-based services and programs in metropolitan and regional areas. These include Adult Counselling Service, Withdrawal Management Service, Stimulant Treatment Program, Opioid Treatment Program, Substance Use in Pregnancy and Parenting Service, Magistrates Early Referral into Treatment Service, Youth Service, and Group Programs (Zhang et al. 2022). Additionally, the service extends its support to hospitals through a specialized Drug and Alcohol Consultation Liaison Service. The consultation liaison staff, comprised of specially trained nurses and doctors, visit patients in hospitals to provide advice and education on the management of substance use–related conditions and facilitate referrals for ongoing support from community-based services.

To date, there has been a lack of research and information on the service utilization and pathways of patients who attended the community-based Drug and Alcohol Service. This knowledge is useful for service providers to identify opportunities to improve efficiency, patient experience, and resource allocation and can serve as a baseline to evaluate improvement initiatives. As alcohol is one of the most commonly abused substances internationally, this study targeted patients with alcohol use disorders and aimed to investigate these patients’ service utilization and pathways in the community-based Drug and Alcohol Service.

Methods

Data source

Data were obtained from the Illawarra Health Information Platform, including three de-identified datasets sourced from ISLHD. These datasets were ED data (December 2011 to January 2019), inpatient data), (October 2011 to January 2019), and community-based Drug and Alcohol Service data (November 2014 to January 2019), which were linked by unique patient identifiers.

Identification of patients with alcohol use disorders

From the ED and inpatient datasets, patients with alcohol use disorders were identified using the Systematized Nomenclature of Medicine Clinical Terms (SNOMED CT) and the International Classification of Diseases, 10th Revision, Australian Modification (ICD-10-AM) (see Appendix S1 for a list of diagnoses used to identify these patients). This led to the identification of 2519 unique patients diagnosed with alcohol use disorders. Among them, 571 patients were found in the dataset of the community-based Drug and Alcohol Service, indicating they used the Drug and Alcohol Service; therefore, they were the target group of this study.

Polysubstance use among patients with alcohol use disorders

Patients with alcohol use disorders may also use another class of substance (e.g. methamphetamine); therefore, their treatment needs, service utilization, and pathway pattern may differ from those who solely use alcohol. To understand the differences between patients with alcohol user disorders who use polysubstance and those who do not, each patient’s polysubstance use status was determined using the SNOMED CT and ICD-10-AM codes listed in Appendix S2. Also, if a patient with alcohol use disorders had interactions with the Opioid Treatment Program or Stimulant Treatment Program provided by the Drug and Alcohol Service, this patient was identified as using polysubstance. After examining all the 571 patients, a total of 87 (15.2%) patients were found to use polysubstance. These patients’ service utilization and pathway patterns were compared with their counterparts who solely used alcohol.

Drug and Alcohol Service encounters and activity

In the Drug and Alcohol Service, a patient encounter starts with the first patient contact and ends when the patient is discharged. A patient may have multiple encounters with the Drug and Alcohol Service over the years; however, the patient can only have one open encounter at any time point.

During an encounter, the patient may go through Intake and Assessment and have multiple contacts with treatment services or programs (e.g. Withdrawal Management Service, Adult Counselling Service, and Opioid Treatment Program) (Qian et al. 2020). The patient’s interaction with the service is a direct care activity (i.e. an activity involving patient contact). The 571 patients incurred a total of 1447 service encounters and 13 974 direct care activities with the Drug and Alcohol Service. These were used to depict the actual patient pathway and derive pathway patterns.

The Drug and Alcohol Service staff also conducted 6104 indirect care activities (i.e. activities with no patient contact, e.g. case planning) for these patients. As this study investigated the service utilization and pathways from the patient’s perspective, indirect care activities were excluded from the analysis.

Data analysis

A descriptive analysis was conducted to describe patient demographics, Drug and Alcohol Service encounters, and direct care activities. The actual patient pathways that occurred in the encounters were depicted with the first three direct care activities using a diagramming and charting tool, Mermaid (Sveidqvist). Patterns were derived from the actual pathways using a process mining tool, Disco (De Roock and Martin 2022). Patients with alcohol use disorders were divided into two groups—patients who use polysubstance and those who do not—for comparison. A Z-test was conducted to compare the direct care activities between these two groups, with a significance level set at P < .05. Pathway patterns of these two groups were also compared.

Results

Patient demographics

From November 2014 to January 2019, 571 patients with alcohol use disorders had direct contact with the Drug and Alcohol Service. Of these patients, 383 (67.1%) were male, 58 (10.2%) were Aboriginal and/or Torres Strait Islanders, and 87 (15.2%) used polysubstance. Thirty-four (6.0%) patients were aged 18–24, 500 (87.5%) were aged 25–64, and 37 (6.5%) were aged 65 or above.

Drug and Alcohol Service encounters and direct care activities

In total, the patients with alcohol use disorders incurred 1447 service encounters and 13 974 direct care activities (i.e. direct patient involvement) in the Drug and Alcohol Service. Intake accounted for 9.7% of all the direct care activities, occurred in 93.2% of the encounters, and was completed for 92.3% of the patients with alcohol use disorders (see Table 1). Assessment accounted for 5.0% of all the activities, occurred in 45.7% of the encounters, and was completed for 73.9% of the patients. Withdrawal Management Service and Adult Counselling Service were the most frequently used services, accounting for 32.7% and 3.6% of the direct care activities, respectively.

Table 1.

Number and percentage of direct care activities, encounters involving each activity type and patients with alcohol use disorders who were provided with this activity in the Drug and Alcohol Service.

Drug and Alcohol Service direct care activity type No. and % of activities
(N = 13,974)		 No. and % of encounters involving this activity
(N = 1447)		 No. and % of patients who were provided with this activity
(N = 571)
Intake 1349 (9.7%) 1349 (93.2%) 527 (92.3%)
Assessment 692 (5.0%) 662 (45.7%) 422 (73.9%)
Withdrawal Management Service activity 4564 (32.7%) 682 (47.1%) 395 (69.2%)
Adult Counselling Service activity 4273 (30.6%) 653 (45.1%) 380 (66.5%)
Opioid Treatment Program activity 1973 (14.1%) 58 (4.0%) 39 (6.8%)
Youth Service activity 305 (2.2%) 32 (2.2%) 20 (3.5%)
Group Programs activity 271 (1.9%) 29 (2.0%) 22 (3.9%)
Stimulant Treatment Program activity 174 (1.2%) 22 (1.5%) 16 (2.8%)
Involuntary Treatment activity 150 (1.1%) 17 (1.2%) 13 (2.3%)
Magistrate Early Referral into Treatment activity 119 (.9%) 21 (1.5%) 20 (3.5%)
Shared Care Service activity 85 (.6%) 3 (.2) 3 (.5%)
Substance Use in Pregnancy and Parenting Services activity 19 (.1%) 6 (.4) 6 (1.1%)
Open in a new tab

Note: The percentage of patients does not add up to 100% because a patient might have more than one activity types. The same is applied to the encounter.

Of all the encounters, 1316 (90.9%) started with Intake, involving 515 (90.2%) patients and 11 531 (82.5%) direct care activities (see Table 2). Among the rest of the encounters, the top three starting activities were Assessment (1.8%), Withdrawal Management Service activity (4.0%), and Adult Counselling Service activity (1.8%).

Table 2.

Number and percentage of encounters that started with a particular type of direct care activity, patients involved in the corresponding encounters, and direct care activities that occurred in these encounters.

Type of starting activity in an encounter No. and % of encounters that started with a particular activity type
(N = 1447)		 No. and % of patients involved in the corresponding encounters
(N = 571)		 No. and % of direct care activities occurred in the corresponding encounters
(N = 13 974)
Intake 1316 (90.9%) 515 (90.2%) 11,531 (82.5%)
Assessment 26 (1.8%) 26 (4.6%) 409 (2.9%)
Withdrawal Management Service activity 58 (4.0%) 57 (10.0%) 658 (4.7%)
Adult Counselling Service activity 26 (1.8%) 26 (4.6%) 307 (2.2%)
Opioid Treatment Program activity 15 (1.0%) 15 (2.6%) 934 (6.7%)
Substance Use in Pregnancy and Parenting Services activity 3 (.2%) 3 (.5%) 29 (.2%)
Youth Service activity 1 (.1%) 1 (.2%) 10 (.1%)
Magistrate Early Referral into Treatment activity 1 (.1%) 1 (.2%) 4 (.0%)
Shared Care Service activity 1 (.1%) 1 (.2%) 92 (.7%)
Open in a new tab

Note: The percentage of patients does not add up to 100% because a patient might have encounters starting with different activity types.

Drug and Alcohol Service encounters that started with Intake

As 90.9% of the encounters started with Intake, investigations into these encounters were conducted. These included (i) comparing the direct care activities that occurred in these encounters between patients with alcohol use disorders who use polysubstance and those who do not; (ii) depicting the actual patient pathways (i.e. the first three activities) in these encounters; (iii) deriving the pathway patterns from these encounters; and (iv) comparing the pathway patterns between patients who use polysubstance and those who do not.

Comparison of direct care activities between patients who use polysubstance and those who do not

A total of 11 531 direct care activities occurred in the encounters that started with Intake (see Table 3). Of these activities, 2303 (20.0%) were incurred by patients who use polysubstance and 9228 (80.0%) by patients who do not use. Compared to patients who do not use polysubstance, patients who use polysubstance had a significantly higher proportion of Intake activities (19.3% vs 11.8%, P < .001), Involuntary Treatment activities (4.0% vs 1.0%, P < .001), and Magistrate Early Referral into Treatment activities (6.4% vs 0.3%, P < .001) but a lower proportion of Withdrawal Management activities (31.2% vs 39.0%, P < .001) and Adult Counselling activities (27.8% vs 37.6%, P < .001).

Table 3.

Comparison of direct care activities occurred in the encounters that started with Intake between patients with alcohol use disorders who also used another drug (i.e. patients who use polysubstance) and those who did not (i.e. patients who do not use polysubstance).

Direct care activities occurred during the encounters starting with Intake No. and % of activities incurred by patients who use polysubstance
(N = 2303)		 No. and % of activities incurred by patients who do not use polysubstance
(N = 9228)		 Total
(N = 11,531)
Intakea 223 (19.3%) 1093 (11.8%) 1316 (11.4%)
Assessment 83 (7.2%) 533 (5.8%) 616 (5.3%)
Withdrawal Management Service activitya 361 (31.2%) 3596 (39.0%) 3957 (39.3%)
Adult Counselling Service activitya 322 (27.8%) 3464 (37.6%) 3786 (37.8%)
Youth Service activity 20 (1.7%) 189 (2.0%) 209 (2.8%)
Group Programs activity 29 (2.5%) 230 (2.5%) 259 (2.2%)
Involuntary Treatment activitya 46 (4.0%) 90 (1.0%) 136 (1.2%)
Magistrate Early Referral into Treatment activitya 74 (6.4%) 30 (.3%) 104 (.9%)
Opioid Treatment Program activity 981 (42.6%) 0 981 (8.5%)
Stimulant Treatment Program activity 162 (7.0%) 0 162 (1.4%)
Substance Use in Pregnancy and Parenting Services activity 0 3 (.03%) 3 (.03%)
Shared Care Service activity 2 (.1%) 0 2 (.02%)
Open in a new tab

Note: Due to no activity or low number of activities, Opioid Treatment Program, Stimulant Treatment Program, Substance Use in Pregnancy and Parenting Services, and Shared Care Service were not compared between the two groups.

a

Statistically significant difference was identified between the two groups.

Actual patient pathways in the encounters that started with Intake

Figure 1 depicts the first three direct care activities that occurred in the actual patient pathways in the encounters that started with Intake. The letter ‘E’ denotes the number and percentage of encounters that included a specific pathway. The letter ‘P’ denotes the number and percentage of patients who experienced a specific path or pathway. The percentage was calculated based on a total of 1447 encounters and 571 patients. For example, the path from ‘Intake- > Assessment’ occurred in 308 (21.3%) encounters and was experienced by 239 (41.9%) patients. The pathway ‘Intake > Assessment > Withdrawal Management’ occurred in 155 (10.7%) encounters and was experienced by 129 (22.6%) patients.

Figure 1.

Figure 1

Open in a new tab

The first three direct care activities of the actual patient pathways in the encounters that started with Intake.

The top five frequent pathways that had at least three activities were (i) pathway ‘Intake > Withdrawal Management > Withdrawal Management’ occurring in 170 (11.7%) encounters and experienced by 130 (22.8%) patients; (ii) pathway ‘Intake > Adult Counselling > Adult Counselling’ occurring in 161 (11.1%) encounters and experienced by 126 (22.1%) patients; (iii) pathway ‘Intake > Assessment > Withdrawal Management’ occurring in 155 (10.7%) encounters and experienced by 129 (22.6%) patients; (iv) pathway ‘Intake > Assessment > Adult Counselling Service’ occurring in 104 (7.2%) encounters and experienced by 93 (16.3%) patients; and (v) pathway ‘Intake > Withdrawal Management > Assessment’ occurring in 86 (5.9%) encounters and experienced by 77 (13.5%) patients. Notably, seven pathways only had two activities (i.e. pathways ending with ‘Nil’). In 313 (21.6%) encounters, Intake was the only activity, and this occurred in 157 (30.5%) patients.

Patterns of patient pathways in the encounters that started with Intake

Figure 2 depicts the patterns derived from the actual patient pathways in the encounters that started with Intake. The circled triangle symbol at the top of the diagram denotes the start of a pathway. The circled square symbol at the bottom of the diagram denotes the end of a pathway. The dashed arrow connects an activity with either of these two symbols, indicating the starting activity (i.e. Intake) or an ending activity (e.g. Group Programs). The box denotes an activity, and the percentage shown in the box is the percentage of encounters that contain this particular activity. For example, 44.2% of the encounters contained Adult Counselling activities. The solid arrow connector denotes a path from one activity to another, and the percentage shown on the path is the percentage of encounters that include this particular path. For example, 13.8% of the encounters included the path ‘Adult Counselling > Withdrawal Management’. The percentage was calculated based on the number of encounters that started with Intake (i.e. 1316). The higher the percentage, the thicker the path and the darker the activity box.

Figure 2.

Figure 2

Open in a new tab

Patterns derived from the actual patient pathways in the encounters that started with Intake.

It can be seen that when patients with alcohol use disorders started their pathways from Intake, their immediate next activities were frequently Withdrawal Management (path: 27.1%), Assessment (path: 24.1%), or Adult Counselling (path: 22.2%). Following Withdrawal Management, the subsequent activities were often Withdrawal Management again (path: 33.9%), Adult Counselling (path: 17.2%), or Assessment (path: 11.2%). Following the Assessment, the most frequent next activity was Withdrawal Management (path: 21.9%). Following Adult Counselling, the most frequent succeeding activities were Adult Counselling (path: 3.7%) and Withdrawal Management (path: 17.2%). A patient pathway could end with any of the activities.

Comparison of pathway patterns between patients who use polysubstance and those who do not

Figure 3 shows the pathway patterns of patients who use polysubstance and those who do not. Different path patterns are marked by an asterisk symbol (*). Compared to those who do not use polysubstance, the path patterns of patients who use polysubstance had Stimulant Treatment Program and Opioid Treatment Program but not Group Programs. Neither did the patients who use polysubstance have the three paths shown in the pathway patterns of those who do not use—the path ‘Withdrawal Management > Assessment’, the path ‘Withdrawal Management > Adult Counselling’, and the path ‘Assessment > Assessment’.

Figure 3.

Figure 3

Open in a new tab

Comparison of pathway patterns between patients who use polysubstance and those who do not in the encounters that started with Intake.

Discussions

To the best of our knowledge, this study was the first to investigate the pathway patterns for patients with alcohol use disorders outside the hospital setting, in the community-based Drug and Alcohol Service. This study explored how patients with alcohol use disorders interacted with Drug and Alcohol Service from November 2014 to January 2019. With 571 patients and a total of 13 974 activities analysed, the impact of polysubstance use in service use and pathway patterns were described. The findings of this paper can provide a baseline for research during the Covid-19 pandemic.

The findings also provide a reference for other Drug and Alcohol Services to improve their services. In the Rojas et al. study, just like in this paper, process mining helped to discover and understand the actual pathways of patients, validate the number of patient cases for each pathway, understand the new pathways, analyse the time performance, interpret the results with the help of experts, and make recommendations for redesigning the process (Rojas et al. 2016).

There were 21.6% of the encounters that stopped at the Intake stage, and the dropout rates were high throughout the patient pathway. This finding followed the previous findings from the hospital setting that dropout rates are highest in the early stages of treatment. A study conducted in a tertiary care de-addiction centre in India showed that alcohol-related dropout rates could be as high as 40% (Basu et al. 2017). Lack of trust may lead to early dropout from hospital-based treatment among patients with alcohol use disorders (Kurata et al. 2023).

Previous research has indicated that a multidisciplinary approach, which includes specialized doctors, is effective in alcohol treatment (Moriarty 2010; Mooney et al. 2014). This multidisciplinary approach expands the circle of support by leveraging individuals who have established positive relationships with the patient, acting as stepping stones (Kurata et al. 2023). Consequently, patients are better positioned to connect with trustworthy professionals. Additionally, patients are more likely to remain in treatment in a hospital setting (Avila et al. 2008). Further research is needed to investigate the reasons behind early treatment dropout in community settings.

Polysubstance abuse is a serious health problem facing society today (Singh 2019). Studies have found that alcohol is often used in conjunction with stimulants such as cocaine (Conway et al. 2013; Crummy et al. 2020; Bonfiglio et al. 2022). However, Stimulant Treatment for patients who use polysubstance with alcohol use disorders in this study only accounted for 1.5% of the total number of encounters, and Opioid Treatment accounted for 4.0%. Jordan et al. (2018) found the strongest association between opioids and acute alcoholism. A study by Baskerville et al. (2023) concluded that alcohol may be used to treat opioid withdrawal symptoms. Polysubstance users may use multiple substances to self-manage symptoms, such as stimulants or alcohol to counteract the depressant effects of opioids (Valente et al. 2020). Alcohol may also alter the pharmacokinetics and pharmacodynamics of co-administered drugs, which means alcohol–drug interactions must be considered when providing alcohol treatment services to patients who use polysubstance (Singh 2019).

Limitations and strength

The length of patient encounters and the number of services included varied widely, making it difficult to generalize to a representative patient pathway. Moreover, Disco software could not reflect the order in which patients received services in this highly repetitive long sequence study. In addition, due to insufficient information identifying the discharge process in the data, each encounter could not be extracted with complete accuracy. This limitation may have impacted the accuracy of the patient pathways generated. In addition, pathways of treatment are often defined by the patient—not the treatment service. Patient preferences may define treatment pathways. Using the process mining technique, this study revealed and visualized the actual treatment pathways and pathway patterns of patients with alcohol use disorders in the community-based Drug and Alcohol Service. The visualized pathways helped service managers to identify issues and opportunities to improve the patient pathways.

Conclusion

This study contributes to an in-depth understanding of service utilization and pathways of patients with alcohol use disorders in the community-based Drug and Alcohol Service. Future studies should investigate these patients’ reasons for dropping out after Intake. Furthermore, this study emphasizes the need to further explore treatment pathways for patients who use polysubstance.

Supplementary Material

Supplement_agae075
supplement_agae075.docx (29.3KB, docx)

Contributor Information

Jingxiang Zhang, Institute of Medical Information, Chinese Academy of Medical Sciences and Peking Union Medical College, Chaoyang District, Beijing 100020, China; Centre for Digital Transformation, School of Computing and Information Technology, Faculty of Engineering and Information Sciences, University of Wollongong, Northfields Ave Wollongong, New South Wales 2522, Australia.

Siyu Qian, Drug and Alcohol Service, Illawarra Shoalhaven Local Health District, 2 Rawson Street Wollongong, New South Wales 2520, Australia.

Guoxin Su, School of Computing and Information Technology, Faculty of Engineering and Information Sciences, University of Wollongong, Northfields Ave Wollongong, Wollongong, New South Wales 2522, Australia.

Chao Deng, School of Medical, Indigenous and Health Sciences, Faculty of Science, Medicine and Health, University of Wollongong, Northfields Ave Wollongong, New South Wales 2522, Australia.

David Reid, Drug and Alcohol Service, Illawarra Shoalhaven Local Health District, 2 Rawson Street Wollongong, New South Wales 2520, Australia.

Barbara Sinclair, Drug and Alcohol Service, Illawarra Shoalhaven Local Health District, 2 Rawson Street Wollongong, New South Wales 2520, Australia.

Ping Yu, Centre for Digital Transformation, School of Computing and Information Technology, Faculty of Engineering and Information Sciences, University of Wollongong, Northfields Ave Wollongong, New South Wales 2522, Australia.

Author contributions

Jingxiang Zhang (CRediT contribution not specified), Siyu Qian (Writing—review & editing [supporting]), Guoxin Su (CRediT contribution not specified), Chao Deng (CRediT contribution not specified), David Reid (CRediT contribution not specified), Barbara Sinclair (CRediT contribution not specified), and Ping Yu (CRediT contribution not specified)

Statements

This paper contains original material, not submitted, in press, or published elsewhere in any form. Each author has contributed significantly to the work and agrees to the submission.

Conflict of interest: None declared.

Ethics approval

Ethics approval was obtained from the joint ISLHD and University of Wollongong Health and Medical Research Ethics Committee. Approval for accessing the data was obtained from the Centre for Health Research Illawarra Shoalhaven Population.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

References

  1. Antonelli D, Bruno G. Application of Process Mining and Semantic Structuring Towards a Lean Healthcare Network. Cham: Springer International Publishing, 2015, 497–508. [Google Scholar]
  2. Arias M, Rojas E, Aguirre S. et al. Mapping the Patient’s journey in healthcare through process mining Int J Environ Res Public Health. 2020;17:6586. 10.3390/ijerph17186586. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Avila P, Marcos M, Avila J. et al. Referral to internal medicine for alcoholism: influence on follow-up care Rev Clin Esp. 2008;208:499–505. [DOI] [PubMed] [Google Scholar]
  4. Baskerville WA, Grodin EN, Lin J. et al. Alcohol Use Among Treatment-Seeking Individuals With Opioid Use Disorder. J Stud Alcohol Drugs. 2023;84:579–84. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Basu D, Ghosh A, Sarkar S. et al. Initial treatment dropout in patients with substance use disorders attending a tertiary care de-addiction Centre in North India Indian J Med Res. 2017;146:S77. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bonfiglio NS, Portoghese I, Renati R. et al. Polysubstance use patterns among outpatients undergoing substance use disorder treatment: a latent class analysis Int J Environ Res Public Health. 2022;19:16759. 10.3390/ijerph192416759. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Butler K, Reeve R, Viney R. et al. Estimating prevalence of drug and alcohol presentations to hospital emergency departments in NSW, Australia: impact of hospital consultation liaison services Public Health Res Pract. 2016;26. 10.17061/phrp2641642. [DOI] [PubMed] [Google Scholar]
  8. Canjels KF, Imkamp MSV, Boymans TAEJ. et al. Improving the Arthrosis Care Process at Maastricht UMC+: Unraveling Complex and Noncomplex Cases by Data and Process Mining . Berlin, Heidelberg: Springer Berlin Heidelberg, 2021. [Google Scholar]
  9. Caron F, Vanthienen J, Vanhaecht K. et al. A process mining-based investigation of adverse events in care processes Health Inf Manag J. 2014;43:16–25. 10.1177/183335831404300103. [DOI] [PubMed] [Google Scholar]
  10. Cho M, Song M, Park J. et al. Process mining-supported emergency room process performance indicators Int J Environ Res Public Health. 2020;17:6290. 10.3390/ijerph17176290. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Conway KP, Vullo GC, Nichter B. et al. Prevalence and patterns of polysubstance use in a nationally representative sample of 10th graders in the United States J Adolesc Health. 2013;52:716–23. 10.1016/j.jadohealth.2012.12.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Crummy EA, O’Neal TJ, Baskin BM. et al. One is not enough: understanding and modeling polysubstance use Front Neurosci. 2020;569-569. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. De Roock E, Martin N. Process mining in healthcare – an updated perspective on the state of the art J Biomed Inform. 2022;127:103995–5. 10.1016/j.jbi.2022.103995. [DOI] [PubMed] [Google Scholar]
  14. Gabrielle dos Santos L, Claudia M, Daniella Yuri M. et al. Process mining leveraging the analysis of patient journey and outcomes: stroke assistance during the Covid-19 pandemic Knowl Manag E-Learn. 2021;13:421–37. [Google Scholar]
  15. Hillestad R, Bigelow J, Bower A. et al. Can electronic medical record systems transform health care? Potential health benefits, savings, and costs Health Aff. 2005;24:1103–17. 10.1377/hlthaff.24.5.1103. [DOI] [PubMed] [Google Scholar]
  16. Illawarra Shoalhaven Local Health District https://www.islhd.health.nsw.gov.au/.
  17. Jordan A, Salen P, Wojda TR. et al. Exploring the association between initial serum alcohol concentration and polysubstance use: more than a simple “gateway drug” effect? Int J Crit Illn Inj Sci. 2018;8:201. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Kim E, Kim S, Song M. et al. Discovery of outpatient care process of a tertiary university hospital using process mining Healthc Inform Res. 2013;19:42–9. 10.4258/hir.2013.19.1.42. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Kurata T, Hashimoto T, Suzuki H. et al. Effect of a multidisciplinary approach on hospital visit continuation in the treatment of patients with alcohol dependence Neuropsychopharmacol Rep. 2023;43:542–52. 10.1002/npr2.12349. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Kwok CS, Muntean E-A, Mallen CD. The patient pathway review: a new method of evaluating clinical practices to understand the complexities of real-world care Crit Pathw Cardiol. 2021;20:213–9. 10.1097/HPC.0000000000000267. [DOI] [PubMed] [Google Scholar]
  21. Kwok CS, Muntean E-A, Foster W. et al. Patient pathways in cardiology: should pharmaceutical and medical device companies care? Crit Pathw Cardiol. 2022;21:57–60. 10.1097/HPC.0000000000000278. [DOI] [PubMed] [Google Scholar]
  22. Lintzeris N, Deacon RM, Shanahan M. et al. Evaluation of an assertive management and integrated Care Service for Frequent Emergency Department Attenders with substance use disorders: the impact project: evaluating an assertive management service for frequent ED attenders with substance use disorders Int J Integr Care. 2020;20:4. 10.5334/ijic.5343. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Metsker O, Kesarev S, Bolgova E. et al. Modelling and Analysis of Complex Patient-Treatment Process Using GraphMiner Toolbox . Cham: Springer International Publishing, 2019. 10.1007/978-3-030-22750-0_65. [DOI] [Google Scholar]
  24. Mooney SR, Horton PA, Trakowski JRJH. et al. Military inpatient residential treatment of substance abuse disorders: the Eisenhower Army Medical Center experience Mil Med. 2014;179:674–8. 10.7205/MILMED-D-13-00308. [DOI] [PubMed] [Google Scholar]
  25. Moriarty KJ. Collaborative liver and psychiatry care in the Royal Bolton Hospital for people with alcohol-related disease. Frontline Gastroenterology. 2010;2:flgastro2105. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Munyisia EN, Reid D, Yu P. Accuracy of outpatient service data for activity-based funding in new South Wales, Australia Health Inf Manag J. 2017;46:78–86. 10.1177/1833358316678957. [DOI] [PubMed] [Google Scholar]
  27. Partington A, Wynn M, Suriadi S. et al. Process Mining for Clinical Processes: a comparative analysis of four Australian hospitals ACM Trans Manag Inf Syst. 2015;5:1–18. 10.1145/2629446. [DOI] [Google Scholar]
  28. Qian S, Munyisia E, Reid D. et al. Trend in data errors after the implementation of an electronic medical record system: a longitudinal study in an Australian regional drug and alcohol service Int J Med Inform. 2020;144:104292–2. 10.1016/j.ijmedinf.2020.104292. [DOI] [PubMed] [Google Scholar]
  29. Reeve R, Arora S, Butler K. et al. Evaluating the impact of hospital based drug and alcohol consultation liaison services J Subst Abus Treat. 2016;68:36–45. 10.1016/j.jsat.2016.05.008. [DOI] [PubMed] [Google Scholar]
  30. Rojas E, Munoz-Gama J, Sepúlveda M. et al. Process mining in healthcare: a literature review J Biomed Inform. 2016;61:224–36. 10.1016/j.jbi.2016.04.007. [DOI] [PubMed] [Google Scholar]
  31. Rovani M, Maggi FM, De Leoni M. et al. Declarative process mining in healthcare Expert Syst Appl. 2015;42:9236–51. 10.1016/j.eswa.2015.07.040. [DOI] [Google Scholar]
  32. Singh AK. Alcohol interaction with cocaine, methamphetamine, opioids, nicotine, cannabis, and γ-hydroxybutyric acid Biomedicine. 2019;7:16. 10.3390/biomedicines7010016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Sun J, Reddy CK. Big data analytics for healthcare. In: Ghani R, Senator TE, Bradley P, Parekh R, He J (eds.) Proceedings of the 19th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. Chicago, Illinois, USA: Association for Computing Machinery, 2013.
  34. Sveidqvist K. Mermaid: Diagramming and charting tool [Online]. Available: https://mermaid.js.org/ (3 November 2024, date last accessed). [Google Scholar]
  35. Valente PK, Bazzi AR, Childs E. et al. Patterns, contexts, and motivations for polysubstance use among people who inject drugs in non-urban settings in the US northeast Int J Drug Policy. 2020;85:102934. 10.1016/j.drugpo.2020.102934. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Van Der Aalst WMP. Process Mining: Data Science in Action. Berlin, Heidelberg: Springer Berlin/Heidelberg, 2016. 10.1007/978-3-662-49851-4. [DOI] [Google Scholar]
  37. Xu H, Pang J, Yang X. et al. Applying Clinical Guidelines to Conformance Checking for Diagnosis and Treatment: A Case Study of Ischemic Stroke . 2020 Online. 2020 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), 2020, 2125–30. [Google Scholar]
  38. Zhang J, Qian S, Su G. et al. Emergency department presentations of patients with alcohol use disorders in an Australian regional health district Subst Abus. 2022;43:1128–40. 10.1080/08897077.2022.2060427. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement_agae075
supplement_agae075.docx (29.3KB, docx)

Articles from Alcohol and Alcoholism (Oxford, Oxfordshire) are provided here courtesy of Oxford University Press

RESOURCES