A Study of Factors Affecting on Patients’ Length of Stay in a Surgical Ward: Toward Optimization by Technology-based Solutions CURRENT

Background: Understanding each of the factors affecting the length of hospitalization especially in surgery wards can play a major role in planning for the optimal use of hospital resources. This study aims to determine factors affecting the length of stay (LOS) in a surgical ward and then provide technology-based solutions . Methods: In this cross-sectional study, 310 records were selected by systematic random sampling from hospitalized patients in surgery ward of a general teaching hospital in Bandar Abbas, situated in sought of Iran. In order to determine the association of 26 variables (demographic, clinical, and non-clinical) with LOS, analytical and descriptive statistics were used. Then, the researchers reviewed relevant literature in PubMed, Scopus, and Google Scholar to introduce solutions based on health information technology (HIT) toward LOS optimization. Results: Mean and median of patients’ LOS in surgery ward were 3.30±3.71 and 2 days respectively. According to multivariate regression model, factors that exerted higher influence on length of stay includes number of para-clinical tests, surgeries, and consultation as well as type of referral and months of admission(p<0.05). Regarding HIT-based intervention, eleven general categories of suggestions were provided. Based on the findings, more accessible technologies such as hospital information system, picture archiving and communication system, telemedicine especially tele-consultation or tele-visit as well as electronic consultation and discharge planning tools alongside smart dashboards for institutions like the center under study in order to expedite the process of diagnosis and treatment, then optimizing LOS seem appropriate. Conclusions: It is important to move toward optimized LOS though understanding and control influential factors; standardize LOS along with continuous monitoring of performance indicators may help to utilize hospital resources more efficiently. HIT-based interventions may support health care providers and administrators to manage patients` admission, hospitalization, transfer, and discharge processes more properly.


Introduction
The increasing cost of healthcare has forced politicians and planners to seek new solutions for cost control and proper usage of limited resources [1].
In hospitals, certain indicators such as bed occupancy rate, average length of stay, bed turnover rate, bed turnover interval, and mortality rate are among the most significant performance indicators which should be calculated and monitored [2,3].
length of stay (LOS) is one of the most important hospital indicators which is required to be checked regularly. It is defined as interval between admission and discharge of a hospitalized patient is an indicator which is commonly used for purposes such as management and planning of hospital care, quality control, and demand of using hospital services. Also, it is an indirect indicator of resource consumption and efficient management of hospital beds; hence, it is an efficiency indicator of hospital performance [4,5].
The patients' longer or shorter than necessary length of stay will influence cost and quality of provided care. In the first case, longer LOS may cause limited resources usages, lower level of service provision to higher number of people, higher pressure for more investment in new treatment centers, lower efficiency and higher depreciation of hospital facilities and more specifically exposure to hospital infection, complications of re-admission and reduction of available resources for patients with critical conditions. On the other hand, shorter-than-necessary length of stay will affect quality of service subversively and contribute to undesirable consequences [4,[6][7][8].
Limitation in treatment centers, personnel, equipment, and increasing costs of healthcare services, may cause more attention to optimize length of stay and related influential factors [9].
Previous studies suggest that a lot of factors influence length of stay. Depending on main goal and studied population, these factors will be different [6,10].
Each of previous studies points to a set of variables as influential upon prediction of prolonged LOS [11,12]. There is no consensus regarding factors affecting patients' length of stay [11], particularly in surgery ward. It is important to study the length of stay and its associated factors in each treatment center individually. Moreover, no study has provided technology-based solutions comprehensively in line with optimizing the patient's length of stay. Thus, this study addresses patients' length of stay and associated factors in general surgery ward as representative of surgical wards of Shahid_Mohammadi teaching hospital in Bandar Abbas; furthermore, it is aimed to provide solutions based on health information technology (HIT) toward LOS optimization.

Methods
This study was conducted in two main steps as follows: The first step was a descriptive-analytical study with a cross-sectional design. The statistical population consists of patients hospitalized in general surgery ward of Shahid_Mohammadi teaching general hospital in Bandar Abbas situated in southern of Iran from March 2016 to March 2017. 310 records were selected from list of hospitalized patients by adopting systematic random sampling. The required information was collected through a checklist that was prepared after review of literature; then, were filled in department of medical records based on Hospital Information System (HIS) fields.
In order to determine the association of 26 variables with patients' length of stay, one-way ANOVA (followed by Least Significant Difference (LSD) post-hoc test), independent sample t-test, Pearson correlation and linear regression were used. The significance level of tests was presumed to be less than 0.05. We used the SPSS software (IBM SPSS Statistics for Windows, Version 23.0. Armonk, NY: IBM Corp; 2015) for data analysis.
In the second step, the researcher reviewed HIT-based solutions to optimize patients' length of stay through a targeted search of available papers in Persian and English languages. To do so, combination of main terms of patient length of stay and information and communication technology (along with their synonyms) were searched in PubMed, Scopus databases as well as Google Scholar without presuming time limits based on a defined search query.
The ethical consideration was approved by the Research Ethics Committee of Hormozgan University of Medical Sciences. Accordingly, the confidentially of information in patients' records was taken into account and the checklists also filled without mentioning patients' names, then the information remained confidential.

Results
The studied general surgical ward had 29 active beds in average during 2016 to 2017; the bed occupancy rate was 74 percent. Based on the results, mean±SD and median of patients' length of stay in this department were 3.30±3.71 and 2 days respectively. The shortest and longest lengths of stays were 1 and 30 days respectively. The patients' mean age was 33.94±21.48 years and respective minimum and maximum ages of 1 and 91 years. In general, 80% of patients underwent surgery that considering the nature of the department, most cases were related to general surgery service. Other patients included three groups 1) according to more clinical examination, they didn't need to surgery, 2) canceled operation for any reason, and 3) re-admission due to surgical complications such as surgical site infection. The results of association between study variables and average length of stay are represented in tables 1-4.  The average length of stay shows significant difference in terms of time of admission, type of insurance and different levels of referrals (P<0.001). The patients who visited the hospital in afternoon shift to be admitted in surgery ward had longer average LOS. In regard to the variable "type of insurance", patients' average length of stay was longer when they used Traffic Accident insurance rather than other insurances (P<0.001).
In cases of referral from emergency department, average length of stay was longer than cases of referral from hospital clinic and other places (P<0.001). The cases in which admitting physician was emergency medicine specialist were accompanied by longer patient length of stay than cases in which patients were admitted by ENT specialist, general surgeon, urology or other surgeons(P<0.01).
According to post-hoc test, internal medicine specialties as attending physician were followed by patients' longer length of stay than orthopedics, urology, and ENT specialists as well as eye surgeon(P<0.05). As a result, emergency admitted cases were accompanied by longer length of stay than nonemergency cases. In regard to the variable of reason of encounter, cases of accident-caused admission were followed by longer length of stay than cases of disease and other events(P<0.001).
The LSD test showed that the difference between the LOS and "cause of hospitalization" related to longer LOS in cases of "neoplasm" and "injury, poisoning and certain other consequences of external causes" compared with those who had other cause (p<0.05).
About patient's condition at time of discharge, in average, death cases had longer length of stay compared with other condition. (P<0.01). was followed by longer length of stay and this association was statistically significant.

As correlation coefficients in table 4 suggest, association of visits and number of para-clinical
services with length of stay was stronger.

Discussion
In this study, the length of stay was found to have a direct and significant association with number Furthermore, results showed that the patients admitted from 2p.m. to 7a.m. had higher length of stay. This might be due to absence of senior specialists during these hours. lack of non-emergency admission during evening and night hours, modification of caring processes, provision of specialized medical and diagnostic services in 24-7 (24 hours a day-7 days a week) manner could prevent from increase of patients' average length of stay to a large extent. Farhadi Hassankiadeh et al revealed that the type of surgery significantly associated with LOS. In patients with appendicitis, hemorrhoids, and skin surgery was a shorter stay [14]. Ravangard et al [15] similarly found out stay due to neoplastic diseases add to patient length of stay. Also in regard to attending physician, they stated that engagement of internal medicine increases length of stay. Current study`s results are in line with these investigations.
It is worth noting, in some cases the cancellation of surgery that leads to longer waiting time in the hospital, was due to lack of patient readiness or failure to perform physician orders, which is mainly due to inadequate training during hospitalization.  -Integrity of clinical and health-related data, improve clinical workflow and safety.
-Observation of patient's medical records with more complete and readable documentation in real-time [18] -Provide alarms and reminders by embedded rule based CDSS: -Prevent from repetition of diagnostic tests.
-Timely and correct prescription of drugs, identification of problems related to drugs side-effects and interventions, reduction of medication error, and better consideration of clinical instructions [22] √ √ Clinical Decision Support System (CDSS) -CDSS based on treatment instructions, protocols. and caring standards could improve and facilitate diagnosis and treatment processes by recommending a series of carerelated processes such as doing tests or prescribing drugs to patients with typical and uncomplicated diagnoses (for example administration of antibiotics and heparin in surgical patients [23]).
-Admission scheduling [24] and discharge planning based on individual patients.

√ √ √
Radio-frequency Identification (RFID) and Barcode Technology -Tracing of patients' location and monitoring of patient turnover [25].
-Exact monitoring of discharging time and facilitated identification of empty beds [26].
-Drug safety and facilitation of drug therapy by confidence in targeting correct patients, correct dose of intended drug by cross-checking [27].
-Identification and warning related to mismatch, drug-related errors, and overdose by comparison of patient's identity and intended drug dose.
-Improve the quality of caring, prevent from medical errors and enable more efficient use of proper resources.
-Reduce the workload of nurses and eliminate unnecessary processes and steps of caring patients [28] √ √ √ Monitoring Systems, Wireless Sensors, and Wearable Tools -Timely and early detecting of the clinical condition and vital signs changes and abnormalities [29,30], especially in highrisk patients.
-Peri/intra-operative management, analysis of post-surgical undesirable infections and side-effects.
-Prevents from reinforcement of unpredicted consequences related to treatment activities or surgery.
√ √ Telemedicine -A lot of potentials for reducing patients' length of stay by this technology [31].
-Tele-pathology [34], tele-radiology [35] and picture archiving and communication system (PACS) for forwarding pathologic and radiologic images to specialists and reception of their viewpoints will help in shortening diagnostic and therapeutic process [35,36]. -Remote monitoring and post-discharge follow-up of surgical patients [44,45] based on patients' condition (especially highrisk groups such as diabetic or elderly patients, also people with immunodeficiency) -Early identification of post-discharge surgical complications, quicker treatment in outpatient centers and reduced rehospitalization and undergoing additional financial expenses.
-Timely detection of unwanted changes in patients clinical conditions can lead to a reduction in LOS of patients who have been re-admitted [46] √ √ Actually at health care institutions, more accessible technologies for LOS optimization are recommended. Among which HIS (by increasing the focus on recording more clinical data), PACS, telemedicine especially tele-consultation or tele-visit as well as electronic consultation and discharge planning tools alongside smart dashboards in order to expedite the process of diagnosis and treatment then optimize the LOS seems appropriate. The role and relationship of the ICT oriented solutions for LOS optimization are presented in figure 1.

Author's contributions
TB, MH and MSH contributed for the research design, data acquisition and analysis. TB, ShRNK and MGH contributed for the manuscript writing. ShRNK also contributed in editing of the manuscript. All authors gave approval for the final version of the manuscript.

Funding
Funding for this study was supported by the Deputy of Research and Technology of Hormozgan University of Medical Sciences, Bandar Abbas, Iran (grant number: 9520).

Availability of data and materials
The dataset used and analysed during the current study are available from the corresponding author on request.

Ethics approval and consent to participate
The ethical consideration was approved by the Research Ethics Committee of Hormozgan University of Medical Sciences, Bandar Abbas, Iran (ethical code: HUMS.REC.1395.56).
We just extracted information from HIS, then confidentiality was maintained based on the rules and policies of the center under study and this was approved by the ethics committee as stated above.
So, didn't need to patients' written consent.

Consent for publication
As our manuscript does not include any individual data or sensitive personal information, therefore consent for publication is "Not Applicable" in this case.

Declaration of conflicting interests
The authors declare that there is no conflict of interest.

Figure 1
The Role of ICT in Optimizing LOS