SCIENTIFIC PAPERS 11.11.2021 FinJeHeW 2021;13(4) 403 Impact of a digital care logistics system on care duration, consumer satisfaction and shift leaders' workload in emergency departments Desale Tewelde Kahsay1, Sanna Salanterä2,3, Janne Engblom4,5, Mikko Häikiö6, Laura-Maria Peltonen2 1 Department of Anaesthesiology, Intensive Care, Emergency Care and Pain Medicine, University of Tur- ku, Turku, Finland; 2 Department of Nursing Science, University of Turku, Turku, Finland; 3 Turku Univer- sity Hospital, Turku, Finland; 4 Department of Mathematics and Statistics, University of Turku, Turku, Finland; 5 School of Economics University of Turku, Turku, Finland; 6 Hospital District of Lapland, Rovaniemi, Finland Laura-Maria Peltonen, Department of Nursing Science, University of Turku, FI-20014 University of Turku, FINLAND. Email: lmemur@utu.fi Abstract The primary goal of introducing digital information systems in healthcare organisations is to improve care processes and outcomes, however, studies that investigate the impact of digital information sys- tems on the day-to-day operations management from the perspective of workflow and consumer satis- faction in emergency departments are scarce. Therefore, this study aimed to explore the impact of a digital clinical logistics system on the duration of patient care, consumer satisfaction and shift leaders' experience of workload in emergency departments. A longitudinal prospective design was used. Three units participated in the study; an intervention unit, a control unit A (no implemented system) and a control unit B (system already in use). We collected data on care duration, consumer satisfaction and shift leaders' experience of workload for four weeks at five time points both before system implementation (summer 2015, spring 2016) and after system imple- mentation (summer 2016, autumn 2016, winter 2016). The average care duration time increased in the postimplementation period in the intervention and control B units (p < 0.001). Duration of care was higher in the intervention unit than control unit B in summer 2016 (p < 0.001) and winter 2016 (p = 0.009). Similarly, duration of care in control unit A was higher than control unit B in spring 2016 (p < 0.001). Consumer satisfaction decreased in the interven- tion unit, in winter 2016 (p < 0.001) and the experience of workload increased in the intervention unit, in summer 2016 and autumn 2016 (p < 0.05). However, the patients-to-nurses ratio was doubled in the SCIENTIFIC PAPERS 11.11.2021 FinJeHeW 2021;13(4) 404 intervention unit in the last time point postimplementation when compared to the first timepoint, while it remained similar in the control units throughout the study period. This work demonstrated that a digital care logistics system may support in increasing the number of patients treated with the same nursing resources. However, this seems to connect to other outcome variables such as increased care duration, increased experience of workload and decreased consumer satisfaction in some postimplementation time points. Keywords: emergency department, digital information system, duration of care, satisfaction, workload Tiivistelmä Digitaalisten tietojärjestelmien käyttöönoton ensisijainen tavoite terveydenhuollon organisaatioissa on parantaa hoitoprosesseja ja tuloksia, mutta tutkimuksia, joissa mitataan digitaalisten tietojärjestelmien vaikutusta päivittäisen toiminnan johtamiseen työnkulun näkökulmasta, on vielä vähän. Tämän tutki- muksen tarkoituksena oli selvittää digitaalisen hoidon logistiikkajärjestelmän vaikutusta potilaiden hoi- don kestoon, palveluita käyttävien tyytyväisyyteen ja vuorovastaavien raportoimaan työn kuormituk- seen päivystyksessä. Tutkimus tehtiin pitkittäisellä asetelmalla. Kolme yksikköä osallistui tutkimukseen; interventioyksikkö, kontrolliyksikkö A (ei tietojärjestelmän käyttöönottoa) ja kontrolliyksikkö B (tietojärjestelmä jo käytös- sä). Aineistoa kerättiin potilaiden hoidon kestosta, kuluttajien tyytyväisyydestä palveluihin, sekä vuoro- vastaavien arvioimasta yksikön työkuormasta neljän viikon intervalleissa yhteensä viisi kertaa - ennen tietojärjestelmän käyttöönottoa (kesä 2015, kevät 2016) ja käyttöönoton jälkeen (kesä 2016, syksy 2016 ja talvi 2016). Potilaiden keskimääräinen hoidon kesto kasvoi käyttöönoton jälkeisenä aikana interventio- ja kontrolli B -yksiköissä (p <0,001). Hoidon kesto oli interventioyksikössä pidempi kuin kontrolli B -yksikössä kesällä 2016 (p <0,001) ja talvella 2016 (p = 0,009). Vastaavasti hoidon kesto kontrolli A -yksikössä oli pidempi verrattuna kontrolli B -yksikköön keväällä 2016 (p <0,001). Kuluttajien tyytyväisyys laski interventioyksi- kössä talvella 2016 (p <0,001) ja heidän raportoima kokemus työkuormasta kasvoi kesällä 2016 ja syksyl- lä 2016 (p <0,05). Potilas per sairaanhoitaja -määrä kaksinkertaistui kuitenkin interventioyksikössä vii- meisenä mittausajankohtana verrattuna ensimmäiseen ajankohtaan, kun luku pysyi samanlaisena molemmissa kontrolliyksiköissä koko tutkimuksen ajan. Tämä työ osoitti, että digitaalinen hoidon logistiikkajärjestelmä voi auttaa lisäämään samoilla hoitore- sursseilla hoidettavien potilaiden määrää. Tämä näyttää kuitenkin olevan yhteydessä muihin tulosmuut- tujiin, kuten hoidon keston pidentymiseen, koetun työkuorman lisääntymiseen ja palveluita käyttävien tyytyväisyyden heikkenemiseen joissakin käyttöönoton jälkeisissä ajankohdissa. Avainsanat: päivystysyksikkö, digitaalinen tietojärjestelmä, hoidon kesto, tyytyväisyys, työkuorma SCIENTIFIC PAPERS 11.11.2021 FinJeHeW 2021;13(4) 405 Introduction Emergency departments (EDs) provide acute care with a broad spectrum of illnesses to individuals without prior appointment. This makes the work in an ED prone to constant and sudden changes. Limited resources need to be coordinated effi- ciently to meet individual care needs. Unit manag- ers are generally responsible for running an ED, but in bigger units and beyond office hours, the responsibility for the day-to-day operations man- agement is often delegated on a shift by shift basis to designated members of staff, i.e. shift leaders [1]. Leadership models in EDs vary and depending on the unit size, the nursing profession shift leader role may typically be stand-alone or combined with a triage nurse role, while the model for the physician shift leader varies between organisations from one designated physician shift leader in the ED to distributed leadership by consultants per speciality [2-5]. Effective information management in this complex environment is a precondition for smooth organisational processes and professionals responsible should have all necessary information in an easily obtainable format for optimal manage- rial decision-making. The information needs in the day-to-day operations management concerns items regarding the number and competence of human resources, patients' health problems, planned care, as well as available material re- sources [6-8]. Electronic health record (EHR) systems are central information sources for patient related infor- mation needs in the day-to-day operations man- agement. But clinicians have reported only mod- erate satisfaction with EHRs, although satisfaction regarding some characteristics seem to be on a slight increase [9]. Further, system usability varies by brand and setting [9-10] Previous research has explored the association of EHRs on workflow and quality of care. Some studies report improvements after EHR implementation [11] while others sug- gest that EHR implementation may impact clini- cians’ task allocation and reduce efficiency during and after implementation [12-14] indicating needs for improvements in usability, functionality and workflow optimisation [13]. Research on barriers to and strategies for successful implementation of digital information systems exists [see e.g. 15-19] and it is essential to understand workflow of all users when planning and implementing digital information systems. Information regarding human and material re- sources needed in the day-to-day operations man- agement is usually spread out in different sources. There are separate systems for diverse aspects of human resources (e.g. rosters, knowledge man- agement and special skills of individual staff mem- bers) and information on materials (e.g. nutrition, medication and medical equipment systems). Much research has focused on technologies to support the allocation of workforce [20,21] and workload of patients [22], but yet there are dis- crepancies with the appropriate allocation of re- sources in the ED [23,24]. Scarce research exists on real-time based digital information systems that combine information about patients and their processes as well as human and material resources to improve professionals’ information manage- ment in the day-to-day operations management although routine health information systems allow evidence-based managerial decisions that support healthcare core functions in planning, monitoring, evaluation and quality improvement [25]. Previous studies have explored the decision-making pro- cess, information needs and challenges concerning information management of shift leaders in the acute care setting [7,26]. Inadequate human re- sources have been associated with increased pa- tient mortality [27], decreased patient safety SCIENTIFIC PAPERS 11.11.2021 FinJeHeW 2021;13(4) 406 [28,29], and reduced quality of care [30]. Paying attention to shift leaders’ information manage- ment is vital to ensure safe and efficient care pro- vision; however, previous research has shown that modern digital information systems fail to support shift leaders’ information processing sufficiently [31,32]. The aim of the study was to explore the impact of a digital clinical logistics system on the duration of patient care, consumer satisfaction and shift lead- ers' experience of workload in the ED. The follow- ing research questions guided the research: 1. What is the association before and after the implementation of the clinical logistics system on duration of care between and within the units? 2. What is the association before and after the implementation clinical logistics system on con- sumer satisfaction between and within the units? 3. What is the association before and after the implementation clinical logistics system on ex- perience of workload between and within the units? Materials and methods Design A longitudinal prospective design was used. ED work processes are complex and they cannot be measured with a single data point or by one statis- tical method but need to be explored from a com- bination of data gathered with multiple methods, and therefore, a sociotechnical systems frame- work and a mixed-methods approach have been suggested for implementing larger information technology projects [33]. Also, a longer-term fol- low-up on the effects of system implementation is recommended as a systematic review indicated that EHR system implementation first increased documentation time, but as staff become more familiar with the system it ultimately improved work flow [13]. Hence, data were collected with different methods during five time points - two before the system implementation and three af- ter: • 1. pre-implementation (summer 2015): 1st to June 30, 2015 • 2. pre-implementation (spring 2016): Febru- ary 15 to March 15, 2016 • 1. post-implementation (summer 2016): 1st to June 30, 2016 • 2. post-implementation (autumn 2016): 15 August to 15, September 2016 • 3. post-implementation (winter 2016): No- vember 15 to December 15, 2016 The reason for a longer interval between data collected in time points one and two was that the system implementation was delayed from autumn 2015 to spring 2016. Each data collection time point lasted four weeks. Data were collected on the duration of care, consumer satisfaction and shift leaders' experience of workload in the unit. An ethical statement was obtained from the Ethics Committee of the University of Turku (Ref. 39/2015-45/2015) and administrative approvals were obtained from each hospital district prior to data collection. Intervention The intervention was the Columna Clinical Logis- tics® system [34]. This digital information system has been designed to support care delivery and workflow within the ED by displaying information about staff (e.g. number and profession of staff on duty), patients and their care processes (e.g. the SCIENTIFIC PAPERS 11.11.2021 FinJeHeW 2021;13(4) 407 patients' main health problem, planned and com- pleted interventions) as well as available resources (e.g. beds). Although the system contains infor- mation about patients and their care, it does not function as an EHR, but rather as a workflow and communication tool for professionals about e.g. distribution of work and patients in the unit, the stage of the care process of individual patients and tasks to be completed. This system has adjusted view options for different purposes and it works on large displays in professionals’ workstations and service user waiting halls, personal computers around the organisation and mobile devices. It can integrate with different systems such as hospital information systems and clinicians’ phones. Setting and sample Three EDs (control unit A, intervention unit and control unit B) from three hospital districts partici- pated. Control unit A is in the southern part, the intervention unit is in the northern part, and con- trol unit B is in the middle of Finland. The units were selected purposely for two reasons. First, the units were similar in function concerning the num- ber and speciality of the professionals. Second, the units admit about the same number of patients (40000-60000) annually with similar health prob- lems both to the general and specialised care pathways. The new digital information system was implemented during the study period in the inter- vention unit. The intervention unit was compared to the control units; control unit A which runs without a digital logistics information system, and control unit B in which the system was imple- mented in 2013. Duration of care Duration of patient care provided (in hours) for each patient was extracted from the hospital in- formation systems. The number of patients who received care during each data collection time and the number of nurses actively participated in pa- tients' care, were also obtained from each unit. Consumer satisfaction A questionnaire was developed based on litera- ture, where items associated with consumer satis- faction and information management were col- lected [35-37]. Face validity was assessed by a total of ten researchers including two professors with hospital leadership positions - one in nursing science and the other in acute medicine, 2 post- doctoral researchers, and 6 doctoral candidates in nursing science with leadership positions or clinical work experience in acute care. All patients and escorts from the three EDs who had been admit- ted during the data collection time points and volunteered to respond were targeted. Question- naires were placed in the "waiting hall" of each ED. The questionnaire for satisfaction had five items, including my satisfaction with 1) the visit to the ED, 2) the length of the waiting time, 3) the in- forming about the waiting time, 4) how I was en- countered, and 5) the arrangements for my care in general. Each item was rated on a scale of 1 to 5 (1= poor, 2 = satisfactory, 3 = good, 4 = very good, and 5 = excellent). The mean value was calculated by summing up the items and dividing the number with the total number of items. The minimum average score was 5 and the maximum average score was 25. The internal consistency in this sam- ple was excellent, as the Cronbach’s α for the scale was 0.95 [38]. Workload in the unit Nursing shift leaders manually documented their estimation of the workload in the units every shift. Workload was documented on a rating scale from 0 to 5 (0 = no hurry in unit, 5 = extremely busy in unit). This rating scale has been used in EDs in SCIENTIFIC PAPERS 11.11.2021 FinJeHeW 2021;13(4) 408 other studies [39] where the ED workload has been considered as none; very light, where availa- ble resources exceed the demand (1); light, where the resources somewhat exceed the demand (2); moderate, where demand and resources approxi- mately match (3); heavy, where the demand somewhat exceeds the available resources (4); and overwhelming, where the demand greatly exceeds the available resources (5). Confounding factors The total number of patients cared for and the total number of nurses during each data collection period were extracted from the hospital infor- mation systems. Patients to nurse ratios are pre- sented. The three participating units did not un- dergo significant organisational changes during the study, which may have interfered with the collect- ed data or its interpretation. One researcher from the team shadowed nursing and medical shift leaders for 3-5 days during data collection time points in all units and documented workflow to ensure consistency. Data analysis In the descriptive statistics, indices of central ten- dency and dispersion, such as mean, frequency and standard deviation were presented as appro- priate. Factorial ANOVA [40,41] was used to ex- plore if an interaction existed between means of two or more independent variables across a single dependent variable. One-way ANOVA was com- puted to test the difference of more than two means of one dependent variable, while the Stu- dents T-test was used to examine the difference of two means of a single dependent variable. Pair- wise analysis of variance using Tukey-Kramer ad- justment was done if differences between means were found when performing factorial and one- way ANOVA. Statistical analysis was performed using SAS® version 9.4 and p-values < 0.05 were considered as significant. Results Duration of care An interaction was found between the units and time points when exploring differences in duration of patient care, F = 3.91, P < 0.001. Pairwise analy- sis showed a difference in the duration of care between the units (Figure 1). The intervention unit had a higher duration of care than control unit B both in summer 2016 (95% CI = 0.20 to 1.47: p <0.001) and winter 2016 (95% CI = 0.08 to 1.26; p = 0.009) (Figure1-A and Table 2). Control unit A had significantly higher duration of patient care compared to control unit B in autumn 2016 (95% CI = 0.20 to 1.47; p = < 0.00) (Figure1-A and Table 2). SCIENTIFIC PAPERS 11.11.2021 FinJeHeW 2021;13(4) 409 Table 1. Mean and standard deviation of duration of care, consumer satisfaction and unit workload. P: N = patients to nurses’ ratio, SD = standard deviation, N in duration = number of patients, N in satisfaction = number of con- sumers (Patients and escorts), N in workload = number of shift leaders Time Control unit A Intervention unit Control unit B Duration of care (hours) N P: N Mean SD N P: N Mean SD N P: N Mean SD Summer 2015 4331 55.5 3.77 3.09 1491 46.6 3.65 3.32 3552 57.2 3.35 2.75 Spring 2016 4223 54.1 4.44 3.19 2821 88.2 3.76 3.74 3872 62.5 3.55 5.35 Summer 2016 4576 58.7 4.81 3.13 2796 87.4 5.32 6.09 3739 60.3 4.49 8.35 Autumn 2016 4659 59.7 5.02 3.72 3011 94.1 5.04 7.92 4485 72.3 4.56 9.73 Winter 2016 4613 59.1 5.28 3.79 3158 98.7 5.39 9.43 4533 73.1 4.73 14.59 Consumer satisfaction N Mean SD N Mean SD N Mean SD Summer 2015 10 10.60 5.72 47 12.53 6.09 41 9.61 5.29 Spring 2016 7 8.86 3.76 15 9.73 3.69 62 11.53 6.37 Summer 2016 9 9.33 6.71 21 10.29 4.90 53 13.09 7.48 Autumn 2016 13 12.15 5.21 21 13.81 6.51 30 10.50 5.06 Winter 2016 9 10.56 3.71 35 5.74 1.59 52 10.6 4.55 Unit workload N Mean SD N Mean SD N Mean SD Summer 2015 84 2.85 0.92 34 2.71 1.24 79 2.55 1.23 Spring 2016 69 2.64 0.83 8 3.19 1.49 74 2.51 0.97 Summer 2016 49 3.09 0.64 25 3.44 1.01 74 2.34 0.99 Autumn 2016 94 2.74 0.91 15 3.73 0.88 76 2.83 0.97 Winter 2016 67 3.13 0.86 19 2.95 1.27 56 3.13 0.82 SCIENTIFIC PAPERS 11.11.2021 FinJeHeW 2021;13(4) 410 Table 2. Pairwise analysis of variance between and within units for care duration (in hours). Effect Between units Units Units MD SE 95%CI p-value Summer 2015 Control Unit A Intervention unit 0.12 0.24 -0.67-0.92 1.000 Control Unit A Control unit B 0.42 0.18 -0.21-1.04 0.611 Intervention unit Control unit B 0.29 0.23 -0.49-1.08 0.995 Spring 2016 Control Unit A Intervention unit 0.68 0.21 -0.02-1.39 0.070 Control Unit A Control unit B 0.89 0.19 0.23-1.55 <0.001 Intervention unit Control unit B 0.21 0.18 -0.42-0.84 0.998 Summer 2016 Control Unit A Intervention unit -0.51 0.21 -1.22-0.19 0.469 Control Unit A Control unit B 0.32 0.19 -0.34-0.99 0.950 Intervention unit Control unit B 0.83 0.19 0.20-1.47 <0.001 Autumn 2016 Control Unit A Intervention unit -0.02 0.21 -0.72-0.68 1.000 Control Unit A Control unit B 0.46 0.19 -0.19-1.11 0.517 Intervention unit Control unit B 0.48 0.18 -0.12-1.08 0.283 Winter 2016 Control Unit A Intervention unit -0.12 0.20 -0.81-0.57 1.000 Control Unit A Control unit B 0.55 0.19 -0.09-1.20 0.194 Intervention unit Control unit B 0.67 0.17 0.08-1.26 0.009 Within units Units Units MD SE 95%CI p-value Control Unit A Summer 2015 Spring 2016 -0.67 0.20 -1.36-0.019 0.0665 Summer 2015 Summer 2016 -1.04 0.20 -1.73 - -0.35 <0.001 Summer 2015 Autumn 2016 -1.25 0.21 -1.95 - -0.55 <0.001 Summer 2015 Winter 2016 -1.51 0.20 -2.20 - -0.82 <0.001 Spring 2016 Summer 2016 -0.37 0.22 -1.11-0.37 0.932 Spring 2016 Autumn 2016 -0.58 0.22 -1.32-0.164 0.342 Spring 2016 Winter 2016 -0.84 0.22 -1.58 - -0.10 0.010 Summer 2016 Autumn 2016 -0.21 0.22 -0.95-0.53 0.999 Summer 2016 Winter 2916 -0.47 0.22 -1.21-0.27 0.698 Autumn 2016 Winter 2016 -0.26 0.22 -1.01-0.48 0.998 Intervention Unit Summer 2015 Spring 2016 -0.11 0.24 -0.92-0.70 1.000 Summer 2015 Summer 2016 -1.68 0.24 -2.49 - -0.86 <0.001 Summer 2015 Autumn 2016 -1.39 0.24 -2.19 - -0.59 <0.001 Summer 2015 Winter 2016 -1.75 0.23 -2.55 - -0.96 <0.001 Spring 2016 Summer 2016 -1.57 0.20 -2.24 - -0.89 <0.001 Spring 2016 Autumn 2016 -1.28 0.20 -1.94 - -0.62 <0.001 Spring 2016 Winter 2016 -1.64 0.19 -2.30 - -0.98 <0.001 Summer 2016 Autumn 2016 0.28 0.19 -0.38-0.95 0.983 Summer 2016 Winter 2016 -0.07 0.19 -0.73-0.58 1.000 Autumn 2016 Winter 2016 -0.36 0.19 -1.00-0.29 0.860 Control unit B Summer 2015 Spring 2016 -0.19 0.17 -0.78-0.39 0.9987 Summer 2015 Summer 2016 -1.14 0.17 -1.73- -0.54 <0.001 Summer 2015 Autumn 2016 -1.21 0.17 -1.77 - -0.68 <0.001 Summer 2015 Winter 2016 -1.37 0.17 -1.94 - -0.81 <0.001 Spring 2016 Summer 2016 -0.94 0.17 -1.52 - -0.36 <0.001 Spring 2016 Autumn 2016 -1.01 0.16 -1.57 - -0.46 <0.001 Spring 2016 Winter 2016 -1.18 0.16 -1.73 - -0.63 <0.001 Summer 2016 Autumn 2016 -0.07 0.17 -0.63-0.49 1.000 Summer 2016 Winter 2016 -0.24 0.16 -0.80-0.32 0.983 Autumn 2016 Winter 2016 -0.17 0.16 -0.70-0.36 0.999 MD = Mean Difference, SE = Standard Error, CI = Confidence Interval SCIENTIFIC PAPERS 11.11.2021 FinJeHeW 2021;13(4) 411 Consumer satisfaction Responses from 425 consumer satisfaction ques- tionnaires were analysed. We found an interaction within the three units in the different time points (F = 4.21, p < 0.001). The pairwise analysis re- vealed that satisfaction was lower in the interven- tion unit when compared to control unit B in win- ter 2016 (95% CI = -9.058 to -0.726; p = 0.006) (Figure 2-A and Table 3). There were no differ- ences in satisfaction between the three units in the other time points. When we looked at the difference within the units, satisfaction in the in- tervention unit was lower in winter 2016 com- pared to summer 2015 (95% CI = 2.53 to11.04; p <0.0001) and autumn 2016 (95% CI = 2.81 to 13.33; p <0.0001) (Figure 1.2-B and Table 3). No difference was found within the other units in the different time points. SCIENTIFIC PAPERS 11.11.2021 FinJeHeW 2021;13(4) 412 Table 3. Pairwise analysis of variance between and within units for consumer satisfaction. Time Between units Unit Unit MD SE 95% CI p-value Summer 2015 Control Unit A Intervention unit -1.93 1.94 -8.57-4.70 0.999 Control Unit A Control unit B 0.99 1.97 -5.73-7.71 1.000 Intervention unit Control unit B 2.92 1.19 -1.150-7.00 0.4801 Spring 2016 Control Unit A Intervention unit -0.88 2.56 -9.61-7.85 1.00 Control Unit A Control unit B -2.68 2.23 -10.27-4.92 0.997 Intervention unit Control unit B -1.79 1.61 -7.28-3.68 0.999 Summer 2016 Control Unit A Intervention unit -0.95 2.22 -8.54-6.64 1.000 Control Unit A Control unit B -3.76 2.01 -10.63-3.11 0.868 Intervention unit Control unit B -2.81 1.44 -7.72-2.11 0.826 Autumn 2016 Control Unit A Intervention unit -1.66 1.97 -8.38-5.07 1.000 Control Unit A Control unit B 1.65 1.85 -4.67-7.98 0.999 Intervention unit Control unit B 3.31 1.59 -2.11-8.73 0.7456 Winter 2016 Control Unit A Intervention unit 4.81 2.09 -2.31 -11.93 0.5867 Control Unit A Control unit B -0.08 2.02 -6.96-6.80 1.000 Intervention unit Control unit B -4.89 1.22 -9.06 - -0.73 0.006 Within units Unit Unit MD SE 95% CI p-value Control unit A Summer 2015 Spring 2016 1.74 2.75 -7.65-11.13 1.000 Summer 2015 Summer 2016 1.27 2.57 -7.49-10.02 1.000 Summer 2015 Autumn 2016 -1.55 2.35 -9.57-6.46 1.000 Summer 2015 Winter 2016 0.04 2.57 -8.71-8.80 1.000 Spring 2016 Summer 2016 -0.48 2.81 -10.08-9.13 1.000 Spring 2016 Autumn 2016 -3.30 2.62 -12.23-5.63 0.995 Spring 2016 Winter 2016 -1.70 2.81 -11.30-7.90 1.000 Summer 2016 Autumn 2016 -2.82 2.42 -11.08-5.44 0.9979 Summer 2016 Winter 2016 -1.22 2.63 -10.21-7.76 1.00 Autumn 2016 Winter 2016 1.60 2.42 -6.66-9.86 1.00 Intervention unit Summer 2015 Spring 2016 2.80 1.66 -2.85-8.45 0.93 Summer 2015 Summer 2016 2.25 1.47 -2.76-7.25 0.97 Summer 2015 Autumn 2016 -1.28 1.47 -6.28-3.72 0.99 Summer 2015 Winter 2016 6.79 1.25 2.54-11.04 <0.001 Spring 2016 Summer 2016 -0.55 1.89 -6.99-5.89 1.00 Spring 2016 Autumn 2016 -4.08 1.89 -10.52-2.37 0.69 Spring 2016 Winter 2016 3.99 1.72 -1.89-9.87 0.57 Summer 2016 Autumn 2016 -3.52 1.72 -9.40-2.36 0.77 Summer 2016 Winter 2016 4.54 1.54 -0.72-9.80 0.17 Autumn 2016 Winter 2016 8.07 1.54 2.81-13.33 <0.001 Control unit B Summer 2015 Spring 2016 -1.92 1.12 -5.76-1.91 0.92 Summer 2015 Summer 2016 -3.48 1.16 -7.48-2.47 0.15 Summer 2015 Autumn 2016 -0.89 1.34 -5.47-3.69 1.00 Summer 2015 Winter 2016 -1.02 1.17 -5.01-2.96 0.99 Spring 2016 Summer 2016 -1.56 1.04 -5.13-2.00 0.97 Spring 2016 Autumn 2016 1.03 1.24 -3.21-5.27 1.00 Spring 2016 Winter 2016 0.90 1.05 -2.68-4.48 0.99 Summer 2016 Autumn 2016 2.59 1.28 -1.76-6.95 0.77 Summer 2016 Winter 2016 2.46 1.09 -1.26-6.18 0.62 Autumn 2016 Winter 2016 -0.13 1.28 -4.50-4.23 1.00 MD = Mean Difference, SE = Standard Error, CI = Confidence Interval SCIENTIFIC PAPERS 11.11.2021 FinJeHeW 2021;13(4) 413 Workload in the unit There was no significant interaction between time- points, units and shifts, F = 1.574, P = 0.07. Similar- ly, there was no interaction between time-points and shifts, F= 1.010, P = 0.427, or between units and shifts, F= 1.00, P = 0.407. However, an interac- tion was found between time-points and units, F = 4.1, P < 0.001. A pairwise analysis demonstrated that workload was higher in the intervention unit compared to the control unit B in summer 2016 (95% CI = 0.337 to1.867; p < 0.001) and control unit A in Autumn 2016 (95% CI = 0.069 to 1.908; p < 0.021) (Figure 3-A and table 4). Similarly, work- load in control unit A was higher than control unit B in Summer 2016 (95% CI = 0.145 to 1.362; p = 0.003) (Figure1.3-A and table 4). Differences within each unit in different time- points were also seen. In the intervention unit, workload was higher in autumn 2016 when com- pared to summer 2015 (95 % CI = 0.003 to 2.052; p = 0.049) (Figure 3-B and Table 4). On the other hand, workload in control unit B was lower in win- ter 2016 compared to the summer 2015 (95 % CI =-1.160 to -0.006; P =0.045), the spring 2016 (95 % CI = -1.205 to -0.035; p = 0.026) and the summer 2016 (95 % CI = -1.381 to -0.210; p = 0.001) (Figure 1.3-B and Table 4). Shift had no interaction with the time points and units on workload. However, shift alone had signif- icant effect on the workload (F = 15.76, p < 0.001); the average workload being the highest in evening shift (3.12, SD = 0.876) compared to morning shift (M = 2.77, SD = 0.897) and night shifts (M = 2.52, SD = 1.133), F = 26.63, P < 0.001. SCIENTIFIC PAPERS 11.11.2021 FinJeHeW 2021;13(4) 414 Figure 1. Duration of care, consumer satisfaction and shift leaders’ experience of workload presented between and within units. Figure 1.1.A shows the ratio of patients to nurses as well as the average care duration in hours between units, and Figure 1.1.B, shows the average care duration in hours within each unit. Boxes represent mean, and the middle line (error bars) on each box represents standard deviation. The three lines above the error bars in "A" represent the number of patients cared for by one nurse in each unit in the five-time paints. *indicates significance, p < 0.05. Figure 1.2.A shows the mean and standard deviation of consumers satisfaction between the three units, and Figure 1.2.B shows the mean and standard deviation of consumers satisfaction within each unit in different time-points. Boxes represent mean, and the middle line (error bars) on each box repre- sents standard deviation. *indicates significance, p < 0.05. Figure 1.3.A shows the mean and standard deviation of shift leaders experience of workload between the three units, and Figure 1.3.B shows mean and standard deviation of shift leaders experi- ence of workload within each unit in the different time-paints. Boxes represent mean, and the middle line (error bars) on each box represents standard deviation. *indicates significance, p < 0.05. SCIENTIFIC PAPERS 11.11.2021 FinJeHeW 2021;13(4) 415 Table 4. Pairwise analysis of variance between and within units for workload. Time Between units Unit Unit MD SE 95%CI p-value Summer 2015 Control Unit A Intervention unit 0.14 0.20 -0.533-0.811 1.00 Control Unit A Control unit B 0.29 0.15 -0.2240.813 0.83 Intervention unit Control unit B 0.16 0.20 -0.523-0.833 1.00 Spring 2016 Control Unit A Intervention unit -0.54 0.36 -1.778-0.692 0.97 Control Unit A Control unit B 0.13 0.16 -0.422-0.685 1.00 Intervention unit Control unit B 0.67 0.36 -0.556-1.904 0.87 Summer 2016 Control Unit A Intervention unit -0.35 0.24 -1.161-0.464 0.98 Control Unit A Control unit B 0.75 0.18 0.145-1.363 0.003 Intervention unit Control unit B 1.10 0.22 0.337-1.867 <0.001 Autumn 2016 Control Unit A Intervention unit -0.99 0.27 -1.908--0.070 0.021 Control Unit A Control unit B -0.08 0.15 -0.594-0.426 1.00 Intervention unit Control unit B 0.90 0.27 -0.030-1.838 0.06 Winter 2016 Control Unit A Intervention unit 0.19 0.25 -0.672-1.046 1.00 Control Unit A Control unit B 0.00 0.18 -0.598-0.599 1.00 Intervention unit Control unit B -0.19 0.26 -1.064-0.691 1.00 Within units Unit Unit MD SE 95%CI p-value Control unit A Summer 2015 Spring 2016 0.20 0.16 -0.34-0.74 0.99 Summer 2015 Summer 2016 -0.25 0.17 -0.84-0.35 0.98 Summer 2015 Autumn 2016 0.10 0.15 -0.40-0.60 1.00 Summer 2015 Winter 2016 -0.29 0.16 -0.83-0.25 0.89 Spring 2016 Summer 2016 -0.45 0.18 -1.07.17 0.47 Spring 2016 Autumn 2016 -0.10 0.15 -0.62-0.42 1.00 Spring 2016 Winter 2016 -0.49 0.17 -1.06-0.08 0.18 Summer 2016 Autumn 2016 0.35 0.17 -0.24-0.93 0.78 Summer 2016 Winter 2016 -0.04 0.18 -0.66-0.58 1.00 Autumn 2016 Winter 2016 -0.39 0.16 -0.92-0.14 0.43 Intervention unit Summer 2015 Spring 2016 -0.48 0.38 -1.78-0.82 0.99 Summer 2015 Summer 2016 -0.73 0.26 -1.61-0.14 0.21 Summer 2015 Autumn 2016 -1.03 0.30 -2.05- 0.01 0.049 Summer 2015 Winter 2016 -0.24 0.28 -1.19-0.71 1.00 Spring 2016 Summer 2016 -0.25 0.39 -1.60-1.09 1.00 Spring 2016 Autumn 2016 -0.55 0.43 -1.99-0.90 0.99 Spring 2016 Winter 2016 0.24 0.41 -1.15-1.63 1.00 Summer 2016 Autumn 2016 -0.29 0.32 -1.37-0.79 1.00 Summer 2016 Winter 2016 0.49 0.30 -0.51-1.50 0.94 Autumn 2016 Winter 2016 0.79 0.34 -0.36-1.93 0.55 Control unit B Summer 2015 Spring 2016 0.04 0.16 -0.50-0.57 1.00 Summer 2015 Summer 2016 0.21 0.16 -0.32-0.75 0.99 Summer 2015 Autumn 2016 -0.28 0.16 -0.81-0.25 0.90 Summer 2015 Winter 2016 -0.58 0.17 -1.16-0.01 0.045 Spring 2016 Summer 2016 0.18 0.16 -0.37-0.72 0.99 Spring 2016 Autumn 2016 -0.32 0.16 -0.86-0.22 0.80 Spring 2016 Winter 2016 -0.62 0.17 -1.21- 0.04 0.02 Summer 2016 Autumn 2016 -0.49 0.16 -1.03-0.05 0.12 Summer 2016 Winter 2016 -0.80 0.17 -1.38- 0.21 <0.001 Autumn 2016 Winter 2016 -0.31 0.17 -0.89-0.28 0.90 MD = Mean Difference, SE = Standard Error, CI = Confidence Interval SCIENTIFIC PAPERS 11.11.2021 FinJeHeW 2021;13(4) 416 Discussion The average duration of care increased in the post- implementation period in the intervention unit, and consumers' satisfaction decreased in the in- tervention unit in the last postimplementation time point (winter 2016). Moreover, the workload increased in the intervention unit in two-time points postimplementation. However, the patient- to-nurse ratio was doubled in the intervention unit in the last time point postimplementation when compared to the first, while this number remained relatively similar in both control units. Duration of care Duration of care is considered one of the crucial indicators of the quality of care [42]. Prolonged ED stay is often associated with overcrowding, delay in care, dissatisfaction and poor outcomes [43,44]. In the current study, the duration of patients' care in the EDs was in a pattern of continuous increase. Also, the number of patients in the EDs were on the increase; the increase being more significant in the intervention unit, while it remained more sta- ble in the control units. This difference might indi- cate that the intervention unit was able to double the number of admitted patients without an in- crease in the number of nurses during the study time. Contrary to our expectation, care duration was increased in the intervention unit and the control unit B. Because the effectiveness of care is affect- ed both by internal and external factors [45], im- plementing a digital information system alone might not be a solution to reduce the duration of care. Further studies are needed to identify other factors that impact duration of care. One reason for increased duration of care is that care is getting more complicated [46]. Previous work has shown that use of digital information systems can change working inappropriately, and as a result, the steps required to accomplish a task may increase [47]. Similarly, working with the hospital information system might slow down the normal flow of care [47], and therefore, increase the workload of care providers [28]. Another point of view is that the 8- month follow-up used in this study was not enough to show all advantages of the implement- ed system, as learning how to use such systems may take a long time for the whole unit [13] and hence future research should extend the follow-up times. Research has shown that more complex organisations have more fragmented workflow, which decreases clinicians’ efficiency, but these effects may be mitigated by better information management [49]. More research is needed to explore if the positive impact of digital information systems is superior in more complex healthcare environments, such as large EDs when compared to more simple environments. Consumers satisfaction Consumer satisfaction showed a pattern of fluctu- ation in the different time-points in all units. How- ever, consumer satisfaction was only reduced in the winter of 2016 in the intervention unit. Limited articles were found related to consumer satisfac- tion after system implementation, and incon- sistent findings were reported. In line with our finding, Meyerhoefer et al. reported a drop in sat- isfaction of obstetrics/gynaecology patients after the implementation of a digital hospital infor- mation system [50]. A study by Wali et al. revealed no difference in satisfaction between patients in the intervention and the control group [51]. In contrast to our finding, Lee et al. and Mysen et al. reported an increased satisfaction level of con- sumers after the implementation of a new EHR system [52,53]. SCIENTIFIC PAPERS 11.11.2021 FinJeHeW 2021;13(4) 417 A probable reason for the low consumer satisfac- tion in the last time point in the intervention unit, is be the substantial increase in the number of patients compared to the number of nurses. This imbalance might have increased the workload and nurses would have less time available to give care to each patient. Previous studies indicate that hospitals with a high patient-to-nurse ratio suffer from an excessive workload, with higher risk for increased mortality rates, burnouts and job dissat- isfaction [54,55]. In contrast, an increase in nurse- to-patient ratio is associated with positive nursing and patient outcomes [56,57]. The intervention unit is located in the northern part of the country where daylight time in winter months is scarce. This might have an impact on the mood of people. Kaldenberg reported that pa- tients admitted to a hospital already feeling de- pressed due to the winter weather are more likely to rate their satisfaction for their hospital stay lower [58]. In contrast, enough natural daylight in hospitals has been associated with increased pa- tients' satisfaction, decreased depression, im- proved sleep, as well as decreased hospital stay [59-61]. Unfortunately, we do not have data from winter 2015, which would have helped us to com- pare the effect of winter in different locations. Nonetheless, an average higher satisfaction level in summer 2015 (Table 1) in the same unit seems to support this argument. Lapland, where the intervention unit is located, emerges in winter as a Finnish destination for tourists [62]. Christmas is a special time, as many tourists from different countries with different cultures flock to the arctic town to meet Santa Claus and experience the joyous season in ex- traordinary snowy surroundings [62,63]. Statistic shows that 21% of the tourists in Lapland in winter 2019 were foreigners, while they were only 5% in summer [64]. This increase in the number of tour- ists might contribute to the increased ratio of pa- tients-to-nurses in the intervention unit in the winter season compared to the summer season. This might decrease consumer satisfaction in a crowded ED, particularly for those hospitalised with a different culture and language. A previous study indicated that treatment by foreign nurses is negatively associated with satisfaction regarding communication and overall perception of care [65]. Interventions to improve consumers infor- mation regarding their wait time in the ED are needed as patients have reported having access to wait time information positively impacts on their overall satisfaction with care in the ED [12]. Workload in the unit Workload increased in the intervention unit rela- tive to the control units during the first months of system implementation. This increase can be part- ly explained by the substantial increase of pa- tients-to-nurses ratio in the intervention unit. Be- sides, in the first months of system implementation, professionals were probably overwhelmed with learning a new system, as well as fulfilling their duty to care for patients. In the last time point, the workload returned to its pre- implementation level. Indicating that professionals probably were able to learn the new system and system benefits started showing. This finding is in line with one study that reported a productivity loss of 20% in the first month, 10% a second month and 5% in the third month of sys- tem implementation, with subsequent restoration of productivity to its original level [66]. Besides, another study estimated that each professional spent an average of 134.2 nonclinical hours relat- ed to implementation activities including learning a new digital information system [67]. Implemen- tation of a digital information system might cause SCIENTIFIC PAPERS 11.11.2021 FinJeHeW 2021;13(4) 418 temporary disruption in workflow and loss of productivity while the end-users learn the new technology [68]. In turn, heavy workload and workflow interruptions have been associated with negative patient outcomes, including medication errors, urinary tract infections and fall, mainly during implementation of a new system [69]. In contrast to the reported loss of productivity and increased workload during and after initial imple- mentation, reduced documentation errors and improved safety of patients have been cited in many previous studies, especially months after the implementation of an electronic medical record [42,70,71] For this reason, an increased number of professionals during the implementation and first post-implementation months might reduce the negative consequences of introducing a new sys- tem. But, effective and efficient assignment of clinicians to deliver the desired care has many challenges [72]. Shift work might influence the psychological and physical well-being of the clini- cians that might affect the care outcomes [73]. We found that shift had no interaction with the differ- ent seasons and units to affect the workload; however, in general, workload was highest in the afternoon, followed by the morning shift (Figure 3). Even though a direct study that assessed the effect of shift on the workload was not found, one study reported lower scores of patient-doctor interactions in the afternoon shift, which suggest- ed an increased workload for clinicians in the af- ternoon shift [74]. Limitations This type of design might be susceptible to season variability as some health-related outcomes are known to have a seasonal pattern. Unfortunately, we had an uneven distribution of months before and after implementation of the digital infor- mation system, with winter months missing before the implementation even though it was included after the implementation. We were unable to col- lect data on waiting time to receive treatment and transfer delay after patients complete their treat- ment in the ED, which are vital information to es- timate patients' flow, overcrowding and providers workload. When the ratio of physician-to-patient increases, patient flow in the ED increases; this, in turn, increases patients' satisfaction and decreases waiting time and workload. However, the ratio of physician to patients was not collected, which might act as a confounder to affect the result. Similarly, response rates for satisfaction and expe- rience of workload were low, which may influence the soundness of the findings. Conclusions Our findings indicated that the intervention unit was able to double the number of admitted pa- tients without an increase in the number of nurses during the study period. However, the duration of care and workload increased. Similarly, the satis- faction of consumers reduced in the last data col- lection time. Hence, it is crucial to ensure an ade- quate number of professionals during the implementation of a new digital information sys- tem, to decrease excessive work stress of nurses and increase consumer satisfaction. This study has shown the complexity in measuring the impact of information management in organising care on unit level in EDs. There is a clear need to further explore means of measuring the effectiveness of information management on professionals’ work and patient outcomes. Acknowledgments This study was supported by the Finnish Work Environment Fund (Ref.114249) and the The Finn- ish Nursing Education Foundation. The funding SCIENTIFIC PAPERS 11.11.2021 FinJeHeW 2021;13(4) 419 bodies had no role in designing, collecting, analys- ing, interpreting, and writing of the manuscript. Authors wish to thank the coordinating nurses and staff from the participating hospitals. Conflict of interest The authors declared no conflicts of interest. References [1] Peltonen LM, Lundgrén-Laine H, Siirala E, Löyt- tyniemi E, Aantaa R, Salanterä S. Assessing mana- gerial information needs: Modification and evalua- tion of the Hospital Shift Leaders' Information Needs Questionnaire. J Nurs Manag. 2018 Mar;26(2):108-119. https://doi.org/10.1111/jonm.12515 [2] Sarcevic A, Marsic I, Waterhouse LJ, Stockwell DC, Burd RS. Leadership structures in emergency care settings: a study of two trauma centers. Int J Med Inform. 2011 Apr;80(4):227-38. Erratum in: Int J Med Inform. 2011 Aug;80(8):e48. https://doi.org/10.1016/j.ijmedinf.2011.01.004 [3] Burström L, Nordberg M, Ornung G, Castrén M, Wiklund T, Engström ML, Enlund M. Physician-led team triage based on lean principles may be supe- rior for efficiency and quality? A comparison of three emergency departments with different tri- age models. Scand J Trauma Resusc Emerg Med. 2012 Aug 20;20:57. https://doi.org/10.1186/1757- 7241-20-57 [4] Kayden S, Anderson PD, Freitas R, Platz E, edi- tors. Emergency Department Leadership and Man- agement. Cambridge University Press; 2014. https://doi.org/10.1017/CBO9781139030557 [5] Connolly M, Jacobs S, Scott K. Clinical leader- ship, structural empowerment and psychological empowerment of registered nurses working in an emergency department. J Nurs Manag. 2018 Oct;26(7):881-887. https://doi.org/10.1111/jonm.12619 [6] Peltonen LM, Siirala E, Junttila K, Lundgrén- Laine H, Vahlberg T, Löyttyniemi E, Aantaa R, Sa- lanterä S. Information needs in day-to-day opera- tions management in hospital units: A cross- sectional national survey. J Nurs Manag. 2019 Mar;27(2):233-244. https://doi.org/10.1111/jonm.12700 [7] Lundgrén-Laine H, Kalafati M, Kontio E, Kauko T, Salanterä S. Crucial information needs of ICU charge nurses in Finland and Greece. Nurs Crit Care. 2013 May;18(3):142-53. https://doi.org/10.1111/nicc.12004. [8] Siirala E, Peltonen LM, Lundgrén-Laine H, Sa- lanterä S, Junttila K. Nurse managers' decision- making in daily unit operation in peri-operative settings: a cross-sectional descriptive study. J Nurs Manag. 2016 Sep;24(6):806-15. https://doi.org/10.1111/jonm.12385 [9] Kaipio J, Lääveri T, Hyppönen H, Vainiomäki S, Reponen J, Kushniruk A, Borycki E, Vänskä J. Usa- bility problems do not heal by themselves: Nation- al survey on physicians' experiences with EHRs in Finland. Int J Med Inform. 2017 Jan;97:266-281. https://doi.org/10.1016/j.ijmedinf.2016.10.010 [10] Kaipio J, Kuusisto A, Hyppönen H, Heponiemi T, Lääveri T. Physicians' and nurses' experiences on EHR usability: Comparison between the profes- sional groups by employment sector and system brand. Int J Med Inform. 2020 Feb;134:104018. https://doi.org/10.1016/j.ijmedinf.2019.104018 SCIENTIFIC PAPERS 11.11.2021 FinJeHeW 2021;13(4) 420 [11] Denton CA, Soni HC, Kannampallil TG, Ser- richio A, Shapiro JS, Traub SJ, Patel VL. Emergency Physicians' Perceived Influence of EHR Use on Clin- ical Workflow and Performance Metrics. Appl Clin Inform. 2018 Jul;9(3):725-733. https://doi.org/10.1055/s-0038-1668553 [12] Calder-Sprackman S, Clapham G, Kandiah T, Choo-Foo J, Aggarwal S, Sweet J, Abdulkarim K, Price C, Thiruganasambandamoorthy V, Kwok ESH. The impact of adoption of an electronic health record on emergency physician work: A time mo- tion study. J Am Coll Emerg Physicians Open. 2021 Feb 4;2(1):e12362. https://doi.org/10.1002/emp2.12362 [13] Baumann LA, Baker J, Elshaug AG. The impact of electronic health record systems on clinical documentation times: A systematic review. Health Policy. 2018 Aug;122(8):827-836. https://doi.org/10.1016/j.healthpol.2018.05.014 [14] Pinevich Y, Clark KJ, Harrison AM, Pickering BW, Herasevich V. Interaction Time with Electronic Health Records: A Systematic Review. Appl Clin Inform. 2021 Aug;12(4):788-799. https://doi.org/10.1055/s-0041-1733909 [15] Farzandipur M, Jeddi FR, Azimi E. Factors Af- fecting Successful Implementation of Hospital In- formation Systems. Acta Inform Med. 2016 Feb;24(1):51-5. https://doi.org/10.5455/aim.2016.24.51-55 [16] Alipour J, Mehdipour Y, Karimi A. Factors Af- fecting Acceptance of Hospital Information Sys- tems in Public Hospitals of Zahedan University of Medical Sciences: A Cross-Sectional Study. J Med Life. Oct-Dec 2019;12(4):403-410. doi: 10.25122/jml-2019-0064. [17] Alipour J, Karimi A, Ebrahimi S, Ansari F, Me- hdipour Y. Success or failure of hospital infor- mation systems of public hospitals affiliated with Zahedan University of Medical Sciences: A cross sectional study in the Southeast of Iran. Int J Med Inform. 2017 Dec;108:49-54. https://doi.org/10.1016/j.ijmedinf.2017.10.005 [18] Ahmadian L, Khajouei R, Nejad SS, Ebrahimza- deh M, Nikkar SE. Prioritizing barriers to successful implementation of hospital information systems. J Med Syst. 2014 Dec;38(12):151. https://doi.org/10.1007/s10916-014-0151-9 [19] Zahabi M, Kaber DB, Swangnetr M. Usability and Safety in Electronic Medical Records Interface Design: A Review of Recent Literature and Guide- line Formulation. Hum Factors. 2015 Aug;57(5):805-34. https://doi.org/10.1177/0018720815576827 [20] Lillehol BM, Lønning K, Andersen MH. Explor- ing nurse managers' perception of using the RA- FAELA system as a management tool in a Norwe- gian hospital setting. Nurs Open. 2017 Dec 14;5(1):77-83. https://doi.org/10.1002/nop2.115 [21] Tursunbayeva A, Bunduchi R, Franco M, Pagliari C. Human resource information systems in health care: a systematic evidence review. J Am Med Inform Assoc. 2017 May 1;24(3):633-654. https://doi.org/10.1093/jamia/ocw141 [22] Wang X, Blumenthal HJ, Hoffman D, Benda N, Kim T, Perry S, Franklin ES, Roth EM, Hettinger AZ, Bisantz AM. Modeling patient-related workload in the emergency department using electronic health record data. Int J Med Inform. 2021 Jun;150:104451. https://doi.org/10.1016/j.ijmedinf.2021.104451 [23] Amritzer MA, Muntlin Å, Berg LM, Göransson KE. Nursing staff ratio and skill mix in Swedish emergency departments: A national cross- sectional benchmark study. J Nurs Manag. 2021 Nov;29(8):2594-2602. https://doi.org/10.1111/jonm.13424 SCIENTIFIC PAPERS 11.11.2021 FinJeHeW 2021;13(4) 421 [24] Porcel-Gálvez AM, Fernández-García E, Raffer- ty AM, Gil-García E, Romero-Sánchez JM, Barrien- tos-Trigo S. Factors That Influence Nurse Staffing Levels in Acute Care Hospital Settings. J Nurs Scholarsh. 2021 Jul;53(4):468-478. https://doi.org/10.1111/jnu.12649 [25] Saigí-Rubió F, Pereyra-Rodríguez JJ, Torrent- Sellens J, Eguia H, Azzopardi-Muscat N, Novillo- Ortiz D. Routine Health Information Systems in the European Context: A Systematic Review of Sys- tematic Reviews. Int J Environ Res Public Health. 2021 Apr 27;18(9):4622. https://doi.org/10.3390/ijerph18094622 [26] Kontio E, Lundgren-Laine H, Kontio J, Korvenranta H, Salanterä S. Information utilization in tactical decision making of middle management health managers. Comput Inform Nurs. 2013 Jan;31(1):9-16. https://doi.org/10.1097/NXN.0b013e318261f192 [27] Liang YW, Tsay SF, Chen WY. Effects of nurse staffing ratios on patient mortality in Taiwan acute care hospitals: a longitudinal study. J Nurs Res. 2012 Mar;20(1):1-7. https://doi.org/10.1097/JNR.0b013e3182466ddc [28] Oliveira RM, Leitao IM, Aguiar LL, Oliveira AC, Gazos DM, Silva LM, Barros AA, Sampaio RL. Evalu- ating the intervening factors in patient safety: focusing on hospital nursing staff. Rev Esc Enferm USP. 2015 Feb;49(1):104-13. https://doi.org/10.1590/S0080- 623420150000100014 [29] McGahan M, Kucharski G, Coyer F. Nurse staffing levels and the incidence of mortality and morbidity in the adult intensive care unit: a litera- ture review. Aust Crit Care. 2012 May;25(2):64-77. https://doi.org/10.1016/j.aucc.2012.03.003 [30] Janhunen K, Kankkunen P, Kvist T. Nursing Staff's Perceptions of Quality of Care for Children in Emergency Departments-High Respect, Low Resources. J Pediatr Nurs. Nov-Dec 2017;37:e10- e15. https://doi.org/10.1016/j.pedn.2017.08.029 [31] Peltonen LM, Junttila K, Salanterä S. Front- Line Physicians' Satisfaction with Information Sys- tems in Hospitals. Stud Health Technol Inform. 2018;247:865-869. [32] Peltonen LM, Junttila K, Salanterä S. Nursing Leaders' Satisfaction with Information Systems in the Day-to-Day Operations Management in Hospi- tal Units. Stud Health Technol Inform. 2018;250:203-207. [33] Fouquet SD, Fitzmaurice L, Chan YR, Palmer EM. Doctors documenting: an ethnographic and informatics approach to understanding attending physician documentation in the pediatric emer- gency department. J Am Med Inform Assoc. 2021 Feb 15;28(2):239-248. https://doi.org/10.1093/jamia/ocaa252 [34] Systematic. Our hospital solutions. [Internet] Systematic; 2021 [cited 2. September 2021]. Avail- able from: https://systematic.com/en- gb/industries/healthcare/ [35] Boudreaux ED, O'Hea EL. Patient satisfaction in the Emergency Department: a review of the literature and implications for practice. J Emerg Med. 2004 Jan;26(1):13-26. https://doi.org/10.1016/j.jemermed.2003.04.003 [36] Dinh MM, Enright N, Walker A, Parameswaran A, Chu M. Determinants of patient satisfaction in an Australian emergency department fast-track setting. Emerg Med J. 2013 Oct;30(10):824-7. https://doi.org/10.1136/emermed-2012-201711 [37] Soares AM, Farhangmehr M. Understanding patient satisfaction in a hospital emergency de- partment. Int Rev Publ Nonprofit Market. 2015 Apr 1;12(1):1-5. https://doi.org/10.1007/s12208-014- 0122-3 SCIENTIFIC PAPERS 11.11.2021 FinJeHeW 2021;13(4) 422 [38] Bland JM, Altman DG. Statistics notes: Cronbach's alpha. BMJ. 1997 Feb 22;314(7080):572. https://doi.org/10.1136/bmj.314.7080.572 [39] Clopton EL, Hyrkäs EK. Modeling emergency department nursing workload in real time: An ex- ploratory study. Int Emerg Nurs. 2020 Jan;48:100793. https://doi.org/10.1016/j.ienj.2019.100793 [40] Cody RP, Smith JK. Applied statistics and the SAS programming language. New York, London: North-Holland Publishing Co.; 1987. [41] Cody R. SAS statistics by example. SAS Insti- tute; 2011. [42] Karaca Z, Wong HS, Mutter RL. Duration of patients’ visits to the hospital emergency depart- ment. BMC Emerg Med. 2012 Dec;12:15. https://doi.org/10.1186/1471-227X-12-15 [43] Mentzoni I, Bogstrand ST, Faiz KW. Emergency department crowding and length of stay before and after an increased catchment area. BMC Health Serv Res. 2019 Jul 22;19(1):506. https://doi.org/10.1186/s12913-019-4342-4 [44] Jarvis PRE. Improving emergency department patient flow. Clin Exp Emerg Med. 2016 Jun; 3(2): 63–68. https://doi.org/10.15441/ceem.16.127 [45] Miro O, Sanchez M, Espinosa G, Coll-Vinent B, Bragulat E, Milla J. Analysis of patient flow in the emergency department and the effect of an exten- sive reorganisation. Emerg Med J. 2003 Mar; 20(2): 143–148. https://doi.org/10.1136/emj.20.2.143 [46] Kuperman GJ, McGowan JJ. Potential unin- tended consequences of health information ex- change. J Gen Intern Med. 2013 Dec; 28(12): 1663–1666. https://doi.org/10.1007/s11606-012- 2313-0 [47] Pirnejad H, Bal R, Shahsavar N. The nature of unintended effects of health information systems concerning patient safety: a systematic review with thematic synthesis. Stud Health Technol In- form. 2010;160(Pt 1):719-23. [48] Menachemi N, Collum TH. Benefits and draw- backs of electronic health record systems. Risk Manag Healthc Policy. 2011;4:47-55. https://doi.org/10.2147/RMHP.S12985 [49] Patel VL, Denton CA, Soni HC, Kannampallil TG, Traub SJ, Shapiro JS. Physician Workflow in Two Distinctive Emergency Departments: An Ob- servational Study. Appl Clin Inform. 2021 Jan;12(1):141-152. https://doi.org/10.1055/s- 0040-1722615 [50] Meyerhoefer CD, Sherer SA, Deily ME, Chou SY, Guo X, Chen J, Sheinberg M, Levick D. Provider and patient satisfaction with the integration of ambulatory and hospital EHR systems. J Am Med Inform Assoc. 2018 Aug 1;25(8):1054-1063. https://doi.org/10.1093/jamia/ocy048 [51] Wali RM, Alqahtani RM, Alharazi SK, Bukhari SA, Quqandi SM. Patient satisfaction with the im- plementation of electronic medical records in the Western Region, Saudi Arabia, 2018. BMC Fam Pract. 2020 Feb 15;21(1):37. https://doi.org/10.1186/s12875-020-1099-0 [52] Lee WW, Alkureishi MA, Ukabiala O, Venable LR, Ngooi SS, Staisiunas DD, Wroblewski KE, Arora VM. Patient Perceptions of Electronic Medical Record Use by Faculty and Resident Physicians: A Mixed Methods Study. J Gen Intern Med. 2016 Nov;31(11):1315-1322. https://doi.org/10.1007/s11606-016-3774-3 [53] Mysen KL, Penprase B, Piscotty R. Patient Satisfaction with Electronic Health Record Use by Primary Care Nurse Practitioners. Comput Inform SCIENTIFIC PAPERS 11.11.2021 FinJeHeW 2021;13(4) 423 Nurs. 2016 Mar;34(3):116-21. https://doi.org/10.1097/CIN.0000000000000218 [54] Aiken LH, Clarke SP, Sloane DM, Sochalski J, Silber JH. Hospital nurse staffing and patient mor- tality, nurse burnout, and job dissatisfaction. JA- MA. 2002 Oct 23-30;288(16):1987-93. https://doi.org/10.1001/jama.288.16.1987 [55] Chen YC, Guo YL, Chin WS, Cheng NY, Ho JJ, Shiao JS. Patient-Nurse Ratio is Related to Nurses' Intention to Leave Their Job through Mediating Factors of Burnout and Job Dissatisfaction. Int J Environ Res Public Health. 2019 Nov 29;16(23):4801. https://doi.org/10.3390/ijerph16234801 [56] Kane RL, Shamliyan T, Mueller C, Duval S, Wilt TJ. Nurse staffing and quality of patient care. Evid Rep Technol Assess (Full Rep). 2007 Mar;(151):1- 115. [57] Shin S, Park JH, Bae SH. Nurse staffing and nurse outcomes: A systematic review and meta- analysis. Nurs Outlook. May-Jun 2018;66(3):273- 282. https://doi.org/10.1016/j.outlook.2017.12.002 [58] Schmidt B. Patient Satisfaction Scores Drop During Winter Months. Patient Safety & Quality Healthcare (PSQH); February 28, 2011 [cited 2.9.2021]. Available from: https://www.psqh.com/news/patient-satisfaction- scores-drop-during-winter-months/ [59] Eldaly K, Zaki N, El-Gizawi L. The associations between daylight sufficiency in hospital wards and patient satisfaction with mental healthcare ser- vices: An egyptian sample. Acta Medica Int. 2016;3(2):101-111. https://doi.org/10.5530/ami.2016.2.22 [60] Joseph A. The Impact of Light on Outcomes in Healthcare Settings. Cent Heal Des. 2006;2:1-12. https://www.healthdesign.org/sites/default/files/ CHD_Issue_Paper2.pdf [61] Park MY, Chai CG, Lee HK, Moon H, Noh JS. The Effects of Natural Daylight on Length of Hospi- tal Stay. Environ Health Insights. 2018 Dec 3;12:1178630218812817. https://doi.org/10.1177/1178630218812817 [62] The Nomad Today. Lapland is the favourite destination for foreign tourists. The Nomad Today; February 7, 2019 [Cited 2.9.2021]. Available from: https://www.foreigner.fi/articulo/tourism/lapland -is-the-favorite-destination-of-foreign- tourists/20190207152546001279.html [63] Rimpiläinen G. Critical growth in tourism: Case Rovaniemi. Bachelor Thesis. Degree Programme in Tourism. Rovaniemi: Lapland University of Applied Sciences; 2019 [Cited 2.9.2021]. Available from: http://urn.fi/URN:NBN:fi:amk-2019061016264. [64] Business Finland. Foreign overnights reached a new record in 2019. Business Finland, Visit Fin- land; 2019 [Cited 30.6.2021]. Available from: https://www.businessfinland.fi/4958c7/globalasse ts/julkaisut/visit-finland/tutkimukset/2020/2019- review---travel-in-finland.pdf [65] Mazurenko O, Menachemi N. Use of foreign- educated nurses and patient satisfaction in U.S. hospitals. Health Care Manage Rev. Oct-Dec 2016;41(4):306-15. https://doi.org/10.1097/HMR.0000000000000077 [66] Wang SJ, Middleton B, Prosser LA, Bardon CG, Spurr CD, Carchidi PJ, Kittler AF, Goldszer RC, Fairchild DG, Sussman AJ, Kuperman GJ, Bates DW. A Cost-Benefit Analysis of Electronic Medical Rec- ords in Primary Care. Am J Med. 2003 Apr 1;114(5):397-403. https://doi.org/10.1016/S0002- 9343(03)00057-3 [67] Fleming NS, Culler SD, McCorkle R, Becker ER, Ballard DJ. The financial and nonfinancial costs of SCIENTIFIC PAPERS 11.11.2021 FinJeHeW 2021;13(4) 424 implementing electronic health records in primary care practices. Health Aff (Millwood). 2011 Mar;30(3):481-9. https://doi.org/10.1377/hlthaff.2010.0768 [68] Menachemi N, Collum TH. Risk Management and Healthcare Policy Dovepress Benefits and drawbacks of electronic health record systems. Risk Manag Healthc Policy. 2011;4:47-55. https://doi.org/10.2147/RMHP.S12985 [69] MacPhee M, Dahinten V, Havaei F. The Impact of Heavy Perceived Nurse Workloads on Patient and Nurse Outcomes. Adm Sci. 2017;7(1):731-741. https://doi.org/10.3390/admsci7010007 [70] Devine EB, Hansen RN, Wilson-Norton JL, Law- less NM, Fisk AW, Blough DK, Martin DP, Sullivan SD. The impact of computerized provider order entry on medication errors in a multispecialty group practice. J Am Med Inform Assoc. Jan-Feb 2010;17(1):78-84. https://doi.org/10.1197/jamia.M3285 [71] Khalifa M. Perceived benefits of implementing and using hospital information systems and elec- tronic medical records. Stud Health Technol In- form. 2017;238:165-168. [72] Dall’Ora C, Griffiths P. Shift work in hospitals: what are the effects on patient and employee outcomes? Health Work Evidence Briefs; 2017: 2. University of Southampton; 2017. [73] Verma A, Kishore J, Gusain S. A comparative study of shift work effects and injuries among nurses working in rotating night and day shifts in a Tertiary Care Hospital of North India. Iran J Nurs Midwifery Res. Jan-Feb 2018;23(1):51-56. https://doi.org/10.4103/ijnmr.IJNMR_15_17 [74] Shaker H, Iraji S, Naini AE, Jouibari MAM, Na- sibehvatankhah, Ghavami Y. Effect of shift work on patient-doctor relationship in emergency depart- ment. J Res Med Sci. 2011 Nov;16(11):1495-9.