Revision of hospital work organization using nurse and healthcare assistant workload indicators as decision aid tools

The study was held in a government hospital providing healthcare in medicine, surgery, obstetrics, geriatrics and psychiatry. Due to the merger of several health facilities, its capacity increased from 1679 to 1951 beds in 2011 and 2015, respectively. The number of full-time equivalent of healthcare workers increased from 3429 in 2011 to 3716 in 2015. The average bed occupancy level was over 85%, and more than 215,000 hospital days were recorded per year.

Nurses, healthcare assistants and cleaners, from 23 departments in 2012 that were merged into 20 departments in 2015 (gynaecology, neonatology, paediatrics, cardiac intensive care, cardiology, diabetes, inpatient cardiology, internal medicine, medicine, rheumatology, gastroenterology, orthopaedics, urology, vascular and thoracic surgery, visceral surgery, neurology, oncology, pulmonary department, intensive care unit, short term geriatric assessment), participated in the study. They were asked to provide data on nursing intensity and indirect care.

This was a prospective observational study based on healthcare provider workload recordings in 23 departments from the same institution. It was a descriptive study on nursing intensity. Results were compared for indirect care, patient and healthcare worker satisfaction surveys between 2012 and 2015.

Nursing intensity (SIIPS score) gave an overall and synthetic assessment of daily care required by patients in a given department [7]. The method was based on 278 nursing tasks listed and classified into three categories of care: 1) basic care (feeding, waste disposal, hygiene, clothing, comfort and mobility), 2) technical care (tasks related to medical diagnosis and therapy), 3) relational care (information and support to patients and families). Each care task was timed and classified into 4 groups corresponding to 4 levels of patient dependency, 4 levels of time required for the caregiver to perform these tasks. A time value of eight minutes and 20 s has been set per SIIPS point. Coefficients of 1, 4, 10 and 20 were established based on the intensity of care workload required for a patient. Nursing intensity per patient over 24 h could vary from three points (basic care = 1 + technical care = 1 + relational care = 1) if the patient was autonomous, to 60 points (basic care = 20 + technical care = 20 + relational care = 20) if the patient was totally dependent. As a result, time spent by caregiver could range from 25 to 500 min per patient per day. Nursing intensity was therefore scored by a global assessment, based on care required for a patient over the last 24 h. Technical and relational care was assessed by Nurses, and basic care was assessed by healthcare assistants, present in the morning before the afternoon shift. The choice of rating (1, 4, 10, 20) was based on three grids (See illustration, Additional file 1) that specify the level of intensity corresponding to care required for a patient in each of the three categories of care (basic care, technical care, relational care). A coefficient of nursing intensity per department was calculated (sum of scores divided by 60 and cumulative working hours). A coefficient between 0.6 and 1.0 was considered as normal workload, greater than 1.0 for the overload and less than 0.6 for the sub-activity.

In addition to nursing intensity, healthcare provider workload was quantified through the evaluation of indirect care. These activities were directly linked to the organizational system of each department and correspond to tasks of catering, logistics, administrative and communication, guidance, training, and research (See form, Additional file 2). It refers to tasks that do not directly affect the patient (e.g., when nurses order medication and make examination appointments, or when healthcare workers order meals, ensure bio-cleaning of the environment near the patient, accompany a patient for an examination in another department, or when cleaners ensure bio-cleaning of the unit). The evaluation per department was carried out over seven successive working days (Fig. 1). The percentage of indirect care activity was calculated by dividing the number of hours spent on indirect care by the number of hours worked by healthcare workers in the unit with all staff present during a week.

The distinction between the two indicators is that nursing intensity includes care that directly affects the patient (e.g., giving a medication) while indirect care includes all tasks that do not directly affect the patient (e.g., ordering medication), the caregiver is not at his bedside.

Patient and worker satisfaction surveys have been developed by the management in consultation with the unions (See documents, Additional files 3 and 4).

The process of the study was guided by the Ottawa model of research use [11].

The primary outcome was the nursing intensity score, mirroring patient care needs, which was measured daily from July 2011 to December 2015. The average score per month and per department was used. The other main outcome was the time spent on indirect care by nurses and healthcare assistants, measured over 7 days in 2012 and in 2015. Indirect outcomes like patient and health providers satisfaction were also surveyed. Means +/− standard deviations or medians and ranges, counts, percentages and confidence intervals were used to describe continuous and categorical variables, respectively. Normal distributions were verified and T-test or Mann-Whitney tests were used to compare continuous data of independent samples where appropriate. Chi-square or Fisher tests were used for categorical variables. An alpha level of 0.05 was used for all statistical tests. XLSTAT 2015 (Addinsoft) was used for statistical analysis.

From July 2011 to May 2012, the median of nursing intensity coefficient was 0.71 with a range from 0.47 to 0.90. Therefore, team composition per department was reassessed in order to balance the workload between all services. Four departments (geriatric short stay care, orthopaedic surgery, oncology and visceral surgery) were reinforced by one healthcare worker each. These units that received additional staff had a chronically high care load compared to other units. The result was a decreased disparity in nursing intensity between departments in 2013, which reach a median of 0.67 [0.54–0.78] in September 2013. In 2014, independently of the study, the health organization policy evolved towards the development of very short-term care. As a result, full inpatient beds were closed or converted to partial hospitalization beds. This is why the inpatient units, which previously housed a proportion of semi-autonomous patients and a proportion of highly dependent patients, have seen an increase in the proportion of dependent patients. These more dependent patients needed a higher level of care resulting in an amplified nursing intensity (Fig. 2). It had created new disparities between departments. In December 2015, the median of nursing intensity coefficient was 0.96 with a range from 0.46 to 1.21. These disparities were lessened by the daily use of the reinforcement unit.

Over the study period, all the departments adopted the nursing intensity scoring and its completion rate was 95%. Through the reinforcement unit, more than 86% of absent staff (training, sickness, maternity leave…) were replaced.

On a short term, i.e. on a daily basis, nursing intensity analysis allowed a rebalancing of human resources in hospitalisation departments by comparing the ratings of patient needs and allocating additional staff to the units with the highest care burden. The nursing intensity indicator became a sort of “standard metre”, a steering tool shared by the management team. Specific and shared vocabulary allowed for comparison between departments. In department with overload situation, scores per care were analysed. If basic care had the highest score, an additional healthcare assistant was assigned as reinforcement in the concerned department. If technical care had the highest score, an additional nurse was assigned.

Medium term, i.e. quarterly and annual, analysis of data allowed mapping of the burden of care and characterizing the various hospitalization units through department and patient profiles. It served as a basis for discussion and was a common steering tool.

In the context of restructuring departments, these data were used to model future (long term) organisations and determine staffing requirements in terms of qualifications and number of staff. Patient profiles and nursing intensity scores characterized a unit, and were used for negotiations in the allocation of resources, thus contributing to cost control.

The 1st evaluation of indirect care was held between November 2011 and May 2012. All nurses, healthcare assistants and cleaners comprehensively filled out the indirect care form. During the campaign, 3927 patients were hospitalized in the departments. Globally, the indirect care represented 6144 h out of the 15,977 h of nurses and healthcare assistants working time (Table 1), i.e. 36.4 and 40.9% of time spent on indirect care by nurses and healthcare assistants, respectively (Fig. 3). Indirect care where hospital staff spent most of their time were cleaning, handover, catering, management of working time, delivery and stretcher, medication management, logistics and linen management, administrative tasks and hygiene. Disparities regarding time spent in management of working time, on the phone, and at administrative tasks related to patient medical record were observed between departments. Moreover, this study interestingly highlighted task slips between healthcare assistants and cleaners.

Based on the indirect care activity results, the main organization revision was implemented in April 2014 and concerned five working areas: 1) time spent on handover, 2) working time management (some tasks performed during the day were transferred during the night), 3) connections between services and the pharmacy (nurses and healthcare assistant are no longer going to the pharmacy, but pharmacy technicians are now coming to the units), 4) housekeeping (creation of cleaner teams independent from care units), 5) food management (reorganization with the catering service, optimization of meal cart organisation, ask patients what they do not like instead of what they like) (Fig. 3).

The most efficient service in handover (i.e. intensive care unit) served as a model in order to define a common procedure. At first, written handover was read together by the newly arrived staff. Then, they met the departing staff for oral transmission and exchanged on specific points. Meetings, where teams from all units were brought together, were avoided. Healthcare managers from each department were in charge of implementing this harmonized model of handover.

In order to enable healthcare providers to focus on their healthcare tasks, pooled resources of new logistics professions were created and systematically used starting from April 2014. These logistics professionals included couriers and storemen (14.9 full time equivalents) who transported and stored materials, supplies, meals, and laboratory specimen, and coordinators (20.3 full time equivalents) who comprised formal healthcare assistants and ensured regular supply of the departments.

One year after the implementation of the new organization, in April 2015, nurses and healthcare assistants filled out again the indirect care activity form. Cleaners did not take part of this campaign as they were not any more part of a specific department. Over the study period, 3738 patients were taken care. Globally, the indirect care represented 6062 h out of the 15,985 h of working time (Table 1), i.e. 33.5 and 43.1% of time spent on indirect care by nurses and healthcare assistants, respectively (Fig. 3). The total indirect care activity ratio was reduced by 8% and increased by 5.5% for nurses and healthcare assistants, respectively. At department level, the average number of hours spent by a nurse on indirect care has significantly decreased from 3.0 (+/− 0.4) hours in 2012 to 2.7 (+/− 0.4) hours in 2015 (p = 0.041). Hence, nurses spared 18 min per working-day (341 h a week for the 20 departments) that were reinvested on patients ‘care. The average number of hours spent by a healthcare worker on indirect care increased slightly from 3.3 (+/− 0.5) in 2012 to 3.4 (+/− 0.5) in 2015, but this was not statistically significant (p = 0.441). This new campaign showed an improvement in time spent on handover. Healthcare providers left less frequently and spent less time outside their unit. However, the catering processes still need revision (Fig. 3).

As the reorganization provided more time for nurses at patient bedside and refocused healthcare workers on their own tasks, patient care should have improved and may have had an effect on patient satisfaction. According to the surveys made in 2012 and 2015, patient satisfaction has been maintained, and even seems to have improved concerning room comfort and information received at discharge (See tables, Additional file 6). Nurse and healthcare assistant perception of working conditions were well-preserved (See tables, Additional file 7). Indicators were experienced as an acknowledgment of healthcare workload by healthcare providers.