Accepted: 4 April 2018

DOI: 10.1111/jonm.12647

ORIGINAL ARTICLE

Use of assessment scales, turnover and job strain in nursing

staff: A study in a Colombian hospital

Eddy Triana‐Palencia1 | Luz M. Cárdenas‐Cárdenas1,2 | Arturo Juárez‐García3 | Julia

Quiroz‐Muysina1 | Alvaro J. Idrovo1,4

1
Fundación Cardiovascular de Colombia,
Floridablanca, Santander, Colombia Abstract
2
Grupo de Investigación Municipio Aim: To evaluate the association between job strain and socio‐demographic charac‐
Saludable, Universidad Pedagógica y
teristics, social support, job insecurity, use of patient assessment scales, and turnover
Tecnológica de Colombia, Tunja, Boyacá,
Colombia of nursing staff in a Colombian hospital.
3
Centro de Investigación Transdisciplinar en Background: Nursing is an occupation with high probability of job strain. Use of pa‐
Psicología (CITPsi), Universidad Autónoma
tient assessment scales and turnover of nursing staff could increase exposure to psy‐
del Estado Morelos, Cuernavaca, México
4
Public Health Department, School of chosocial risk.
Medicine, Universidad Industrial de Method: A cross‐sectional study of 222 nurses was conducted. A survey and the Job
Santander, Bucaramanga, Santander,
Colombia Content Questionnaire were used to obtain data at the individual level and free lists
and institutional records were used at the hospital unit level. The associations of inter‐
Correspondence
Luz Mery Cárdenas‐Cárdenas, Escuela est were evaluated with a logistic regression model with robust variance estimator.
de Enfermería, Universidad Pedagógica y Results: Many nurses (50.9%) nurses reported job strain, which was positively associ‐
Tecnológica de Colombia, Calle 24 N° 5 - 63
Antiguo Hospital San Rafael de Tunja, Tunja- ated with high use of patient assessment scales (OR = 2.73; 95% CI = 1.35–5.51) but
Colombia. negatively associated with social support (OR = 0.89; 95% CI = 0.80–0.98). Turnover
Email: [email protected]
was not statistically associated with job strain.
Conclusion: Job strain among nurses was associated with a high use of patient as‐
sessment scales, but not with turnover of nursing staff.
Implications for Nursing Management: The findings of this study suggest possible
opportunities for managers to improve nursing processes, the work conditions of
nursing staff, and the quality of institutions.

KEYWORDS
job strain, nurses, nursing process, turnover

1 | I NTRO D U C TI O N et al., 2012) and issues related to mental health (Stansfeld & Candy,
2006). In nursing, the model has been used to explore the associ‐
According to Karasek’s job demand‐control model (Karasek, 1979), ation between job strain and musculoskeletal diseases, high blood
job strain occurs when workers experience high psychological de‐ pressure, exhaustion, mental health, anxiety, and depression. It has
mands along with low levels of control in the work environment (de‐ also been used to compare job strain among different hospital units
cision autonomy and skill use). This combination of high demands (Juárez‐García, 2007; Juárez‐García et al., 2014; Schmidt, 2013;
and low control is known as “job strain.” The “job strain model” has Trousselard et al., 2016; Zhang, Tao, Ellenbecker, & Liu, 2013).
been the most widely used method for conducting stress‐related re‐ In Latin America, the demand‐control or job strain model has
search and has the power to demonstrate the longitudinal effects of not been widely used in the nursing field. In a literature search of
a job’s environmental and psychosocial conditions on health, partic‐ the MEDLINE and LILACS databases conducted on 6 September
ularly for health outcomes such as cardiovascular disease (Kivimäki 2017 using the strategy “demand control model”[All Fields] OR “job

42  |  wileyonlinelibrary.com/journal/jonm
© 2018 John Wiley & Sons Ltd J Nurs Manag. 2019;27:42–51.
TRIANA‐PALENCIA et al. |
      43

content questionnaire”[All Fields] AND (“nurses”[MeSH Terms] OR experience work overload due to financial policies (Echeverri &
“nurses”[All Fields]), 86 articles were found, only eight of which Vélez, 2005). In addition to providing direct patient care and family
were studies conducted in Latin American countries; seven of those education, nurses also need to manage care units and staff, perform
were in Brazil (Araújo, Aquino, Menezes, Santos, & Aguiar, 2003; secretarial tasks, and handle public relations (Echeverri & Vélez,
Bosi de Souza Magnago et al., 2010; da Silva et al., 2015; Fernandes 2005). As nurses perform many tasks, rather than serving as an im‐
Portela et al., 2015; Henriques da Silva, 2015; Pinho & Araújo, 2007; portant part of the evaluation stage the use of patient assessment
Tavares et al., 2012), and one was in Mexico (Juárez‐García, 2007). scales may become just another requirement that has to be met,
In Colombia, Leguizamón, Gómez Ortiz, Juarez García and others thereby contributing to job strain.
have used the demand‐control model with nurses and other health Lastly, a work environment with high turnover of nursing staff
professionals to study psychosocial factors. Leguizamón’s study has been associated with an increase in work demands due to pro‐
reports that, although nurses in Colombia showed higher demand cesses requiring training, orientation, and supervision of new per‐
scores, they also had higher control scores (Leguizamón & Gómez sonnel (Twigg & McCullough, 2014). As nursing staff may not have
Ortíz, 2002). On the other hand, Juarez‐Garcia’s study found that sufficient control over these demands, they contribute to job strain.
job strain for health professionals in Colombia is higher than in other The overall objective of this study was thus to evaluate the as‐
Latin American countries; however, this study included small sample sociation between job strain and socio‐demographic characteristics,
sizes from only three countries (Chile, Mexico, and Colombia), and social support, job insecurity, use of patient assessment scales, and
their results should not be generalized to all Latin American coun‐ turnover of nursing staff in a Colombian hospital with a high level of
tries (Juárez‐García, Aldo, Viviola, Carlo, & Peter, 2008). complexity. To achieve this overall objective, the following specific
Various studies have identified nurses as an occupational group objectives were established:
with a high psychosocial risk, which contributes to a high probability
of job strain (Fiabane, Giorgi, Sguazzin, & Argentero, 2013; Wu, Chi, • Identify the proportion of nurses with job strain, their socio‐de‐
Chen, Wang, & Jin, 2010). Job strain among nurses has been associ‐ mographic characteristics, and their level of social support and
ated with high blood pressure, burn out, mild psychiatric disorders, job insecurity.
and negative evaluation of one’s own health (Juárez‐García, 2007; • Identify the use of patient assessment scales by the hospital units
Theme Filha, De Souza Costa, & Rodrigues Guilam, 2013; Urbanetto in the institution.
et al., 2013). It has also been shown to affect quality of care and the • Identify turnover of nursing staff in the different hospital units in
administration of hospital services. High job strain among nurses re‐ the institution.
duces work capacity, contributes to errors in care plans, and lowers • Evaluate crude and adjusted associations between job strain and
patient satisfaction (Dewa, Loong, Bonato, Thanh, & Jacobs, 2014; socio‐demographic characteristics, social support, job insecurity,
Fiabane et al., 2013). In the administration of hospital units, strain use of patient assessment scales, and turnover of nursing staff.
among nurses has been associated with work absenteeism due to
illness and a higher risk of resignation (Dewa et al., 2014; Fiabane et
al., 2013; Trybou et al., 2014).
2 | M ATE R I A L S A N D M E TH O DS
Some studies have found job strain to be associated with cer‐
tain socio‐demographic characteristics and work conditions (Tsai
2.1 | Study design and participants
& Liu, 2012; Urbanetto et al., 2011; Wu et al., 2010). The literature
reports that older nurses and those with higher educational levels A cross‐sectional study was performed between October 2012 and
have a lower risk of job strain (Lee & Wang, 2002; Lu, While, & October 2013 at the Cardiovascular Foundation of Colombia. All
Barriball, 2007), whereas nursing staff in the intensive care unit of the 377 nurses who worked in the institution at the time of the
(ICU) and urgent care show higher levels of job strain (Urbanetto study were invited to participate. The response rate was 65%, which
et al., 2011; Zhang et al., 2013). Job insecurity is considered to resulted in a sample of 246 nurses. Those who did not participate
be a significant risk factor for psychosocial strain (Juárez‐García, were on vacation or medical leave, or they were not interested in
2007), and social support from the department head as well as participating. In terms of the hospital unit where the nurses worked,
from colleagues is negatively associated with job strain (Chiu, those with lower response rates were the neonatal ICU (46%), adult
Chung, Wu, & Ho, 2009; Hayes et al., 2012). Nevertheless, the hospitalization (48%), paediatric intermediate ICU (60%), paediatric
evidence for all these associations is inconsistent and no stud‐ cardiovascular ICU (62%), and general paediatric ICU (63%), while
ies have been performed to examine these relationships among transplant ICU (80%) and surgery (79%) had higher response rates.
Colombian nurses. Of the total participating nurses, 145 (59%) were nursing assistants
Patient assessment scales—such as the Fall Risk Assessment, the or certified nursing assistants and 101 (41%) were nursing profes‐
Pain Scale, the Braden Scale, and the Sedation Scale, among oth‐ sionals or registered nurses. The study did not contain the names
ers—are useful for nursing processes during the assessment stage of the participants or identify them individually, and the adminis‐
because they enable nurses to monitor the health status of pa‐ trative and management departments did not have access to that
tients. Nonetheless, nurses in some Colombian health institutions information.
 |
44       TRIANA‐PALENCIA et al.

key nurse informants in each hospital unit were thus asked to list
2.2 | Information
on a piece of paper the patient assessment scales that were used
The study collected information about the nurses and all of the 12 at the hospital unit in which they worked. The significance given to
hospital units at the institution. With regard to the nurses, a survey each one was estimated based on the frequency and order of the
was administered to the participants in order to collect socio‐demo‐ scales mentioned by the key nurse informants (Jerome‐Smith, 1993)
graphic information including gender, age, marital status, educational and the main patient assessment scales used by the nursing staff at
level, position at the institution, months employed, and number of each hospital unit were identified. The key nurse informants who
children. In addition, job characteristics were identified using the Job participated in creating the free lists worked in the following hospital
Content Questionnaire (JCQ). The information about the hospital units: neonatal ICU (n = 16), general paediatric ICU (n = 2), adult car‐
units was obtained from free lists (described below), provided by key diovascular ICU (n = 4), transplant ICU (n = 12), adult hospitalization
informants from each unit (nurses who voluntarily participated in the (n = 23), paediatric cardiovascular ICU (n = 30), adult intermediate
free lists), and from the institution’s administrative records. ICU (n = 4), paediatric intermediate ICU (n = 2), paediatric hospital‐
ization (n = 13) and urgent care (n = 5). Nurses in ambulatory and sur‐
gery units reported that they did not use any of these scales.
2.3 | Outcome
The turnover of the nursing staff was calculated for each of the
Job strain, the combination of high psychological demands and low hospital units in the institution, based on the institutional records
labour control, was the dependent variable in this study (Karasek, corresponding to the four previous years. Turnover was defined as
1979). Job strain among the nurses was evaluated using the JCQ, nursing personnel who left the institution (Hayes et al., 2012). This
designed by Karasek, which was psychometrically validated for the was calculated according to the procedure described by Buchan
Colombian population (Gómez‐Ortíz, 2011). The JCQ version used (2010), in which the numerator represents the number of nurses per
by this study contained 27 Likert items with the following response hospital unit who voluntarily left the institution during the 4 years
choices: “No, completely disagree,” “Disagree,” “Agree,” and “Yes, prior to the study divided by the average number of nurses per hos‐
completely agree.” The dimensions related to psychological de‐ pital unit during the same period of time. Due to lack of information
mands and work control were estimated according to procedures from the institution’s administrative records, turnover of the nursing
described previously by Juarez‐Garcia and others (Cedillo & Karasek, staff did not include nurses who left the institution because of lay‐
2003; Gómez‐Ortíz, 2011; Juárez‐García, 2007). According to the offs, occupational illness, or disability.
demand‐control model, the effect of the synergistic combination
of the dimensions on health is more important than their isolated
2.5 | Statistical methods and analysis of free lists
effects. This study therefore used job strain as its core construct,
which was determined by multiplying psychological demands by two Socio‐demographic characteristics, social support, job insecurity,
and dividing the result by work control (psychological demands × 2/ use of patient assessment scales, and turnover were described using
work control) (Gómez‐Ortíz, 2011; Juárez‐García, 2007). Scores over proportions of categorical variables. In the case of quantitative
1 indicated the presence of job strain. variables, mean and standard deviation or median and interquartile
range were used as measures of central tendency and dispersion ac‐
cording to their normal or non‐normal distribution, respectively.
2.4 | Independent variables
The free lists were processed with Visual Anthropac 4.98 soft‐
The independent variables had two levels: the individual and the ware (Analytic Technologies, Lexington, USA). As an indicator of the
hospital unit. At the individual level, the socio‐demographic charac‐ significance of the use of patient assessment scales, Smith’s salience
teristics included: gender, age, marital status, educational level, po‐ measurements were obtained for each scale and hospital unit. Each
sition at the institution, months employed and number of children. Smith’s salience measurement takes into account the frequency and
Social support and job insecurity were also included at the individual order in which the key nurse informants listed each patient assess‐
level. These were determined based on the JCQ questionnaire and ment scale (Gravlee, 2002; Sinha, 2003). Smith’s salience equals the
calculated using procedures described previously in other publica‐ sum of an item’s percentile rank divided by the total number of lists
tions (Cedillo & Karasek, 2003; Gómez‐Ortíz, 2011; Juárez‐García, (Jerome‐Smith, 1993). The index ranges from 0 to 1, with values near
2007). 1 indicating that the patient assessment scale was mentioned by a
The patient assessment scales and turnover of nursing staff were larger number of key nurse informants and was one of the top scales
evaluated at the hospital unit level. The use of patient assessment on the lists. The average Smith’s salience was calculated for every
scales was evaluated based on free lists (Jerome‐Smith, 1993), a scale listed per hospital unit. In accordance with the logic of free
structured interviewing technique that is useful to explore cultural lists, zero values were imputed for the patient assessment scales
domains (in this study, the use of patient assessment scales for nurs‐ that were not mentioned by participants. The significance of the use
ing processes) at the population‐level (Jerome‐Smith, 1993). When of patient assessment scales was categorized based on the tertile
using free lists, informants are asked about a specific topic and their distribution of the Smith’s salience measurements. A Smith’s salience
responses create lists, which are analysed (Jerome‐Smith, 1993). The less than or equal to 0.153 indicated low use of scales, a salience
TRIANA‐PALENCIA et al. |
      45

over 0.153, and less than or equal to 0.183 moderate use, and over TA B L E 1   Job strain, socio‐demographic characteristics, social
0.183 high use. support and job insecurity among nurses at the Cardiovascular
The crude and adjusted associations between job strain and Foundation of Colombia, 2013 (n = 222)

socio‐demographic characteristics, social support, job insecurity, Characteristic n (%)

use of patient assessment scales and turnover of nursing staff were
Job strain
evaluated using a logistic regression model with robust variance es‐
No 109 (49.1)
timator to have into account observations clustered at the hospital
Yes 113 (50.9)
unit level. A multiple model was adjusted for variables with P < 0.25
Gender
resulting from the crude analysis. The age variable was included in
Female 201 (90.5)
the multiple regression model given the previous results reported in
the literature of the region; for example, Theme Filha et al. (2013) Male 21 (9.5)

found higher job strain at a younger age in Brazilian nurses. Position

Even though 24 nurses (10%) were excluded from the analysis Head of nursing or coordinator 90 (40.5)
due to incomplete information, the characteristics of the included Nursing assistant 132 (59.5)
and excluded subjects were comparable, except for the number Marital status
of children and use of patient assessment scales (data not shown). Single 98 (44.2)
Based on a job strain prevalence of 21% among nurses, which was Married 68 (30.6)
previously reported in the literature (Theme Filha et al., 2013;
Other 56 (25.2)
Urbanetto et al., 2011), the power of the final sample of 222 nurses
Age
was over 85% for associations between job strain and socio‐de‐
≤29 years 77 (34.7)
mographic characteristics previously described in other studies
30–35 years 82 (36.9)
(OR = 2.6 between job strain and under 35 years of age and OR = 2.9
≥36 years 63 (28.4)
between job strain and low social support) (Theme Filha et al., 2013).
A sensitivity analysis was conducted using multiple imputations Educational level

by chained equations for variables with missing data, such as age Technical (nursing assistant) 133 (59.9)

(n = 18) and months employed (n = 7). Twenty imputed datasets and Professional (nursing professional) 62 (27.9)
20 cycles per imputation were used and the analyses were repeated Specialized (nursing professional) 27 (12.2)
with the imputed data. We also ran a multiple regression model with Months employed
robust variance estimator, in which job strain was a continuous vari‐ Less than 60 months 145 (65.3)
able. All of the analyses were performed with the Stata 14 statistical More than 60 months 77 (34.6)
program (Stata Corporation, College Station, TX, USA). Number of children
Median (IQR) 1.0 (2.0)
Social support
3 |  R E S U LT S
Mean (SD) 23.0 (3.9)
Job insecurity
Table 1 shows that 50.9% of the nursing personnel who participated
Mean (SD) 5.1 (1.5)
in this study experienced job strain. Most of the participants in the
sample were women (90.5%), single (44.2%), worked as nursing as‐ SD: Standard deviation; IQR: Interquartile range.
sistants (59.5%), were between 30 to 35 years of age (36.9%), and
had worked at the institution less than 60 months (65.3%) (Table 1). social support, job insecurity, and high use of patient assessment
Table 2 presents the results from the free lists pertaining to the use scales (Table 4). The results for turnover of nursing staff were not
of patient assessment scales for each hospital unit in the institution. significant.
The scales mentioned most frequently were the Pain Scale, the Fall Table 5 shows the results from the logistic regression model with
Risk Scale, the Braden Scale for Predicting Pressure Sore Risk and robust variance estimator adjusted according to individual charac‐
the Nutrition Scale. The units that used the most scales were the teristics as well as hospital unit conditions. These had a P < 0.05 in
general paediatric ICU, the paediatric intermediate ICU, and the the crude analysis, and included educational level, months employed,
adult cardiovascular ICU, as seen in Table 2. social support, job insecurity, and use of patient assessment scales.
The neonatal ICU had the highest turnover of nursing staff of all Job strain among nurses was found to be positively associated with
the hospital units. Equally significant results were also found for the over 60 months of employment (OR = 2.97; 95% CI = 1.74–5.06) and
general paediatric ICU, the ambulatory unit, the adult cardiovascular high use of patient assessment scales (OR = 2.62; 95% CI = 1.34–
ICU, and the transplant ICU (Table 3). The crude logistic regression 5.13). The opposite was found for over 36 years of age (OR = 0.37;
model with robust variance estimator indicated that there was an 95% CI = 0.15–0.91), specialized education level (OR = 0.36; 95%
association between job strain and a specialized educational level, CI = 0.18–0.75), and social support from colleagues and the head
TA B L E 2   Smith's salience for patient assessment scales, by hospital unit. Cardiovascular Foundation of Colombia, 2013
|

Adult Paediatric Adult Paediatric General

46      

Ambulatory Neonatal Adult Intermediate Transplant cardiovascular Paediatric cardiovascular intermediate paediatric Scale
Nursing scales unit Surgery ICU hospitalization Urgency ICU ICU ICU hospitalization ICU ICU ICU mean

Fall risk scale 0.000 0.000 0.000 0.751 0.880 0.875 0.663 0.567 0.846 0.844 1.000 1.000 0.619
Braden scale 0.000 0.000 0.854 0.352 0.523 0.475 0.817 0.498 0.628 0.594 0.700 0.436 0.490
Pain scale 0.000 0.000 0.771 0.836 0.647 0.837 0.633 0.802 0.760 0.750 0.700 0.905 0.637
Neurological 0.000 0.000 0.000 0.190 0.120 0.387 0.525 0.000 0.211 0.313 0.200 0.714 0.222
assessment
scale
Cardiovascular 0.000 0.000 0.000 0.028 0.200 0.000 0.000 0.000 0.000 0.250 0.000 0.200 0.057
risk scale
Nutritional 0.000 0.000 0.375 0.291 0.130 0.000 0.017 0.467 0.390 0.156 0.400 0.809 0.253
assessment
scale
Edmonton 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.063 0.000 0.250 0.026
symptom
assessment
scale
Zung Self‐Rating 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.031 0.000 0.100 0.011
Depression
Scale
Barthel Scale 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.022 0.000 0.000 0.045 0.006
Sedation scale 0.000 0.000 0.000 0.000 0.000 0.050 0.221 0.587 0.066 0.000 0.000 0.318 0.104
Paediatric 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.273 0.023
nutritional
screening
Neonatal 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.227 0.019
nutritional
screening
Nutritional 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.182 0.015
screening of
pregnant
women
Karnofsky 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.150 0.013
Performance
Status Scale
Minnesota Living 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.050 0.004
with Heart
Failure
Questionnaire
(Continues)
TRIANA‐PALENCIA et al.
TRIANA‐PALENCIA et al. |
      47

of nursing (OR = 0.88; 95% CI = 0.80–0.98). Although job insecurity

Smith’s salience varies between 1 and 0, Smith’s salience near to 1 indicated that the patient assessment scale was mentioned by more nurses and was among the top scales on the lists. Zero values were
0.009
mean
Scale
was associated with job strain, this association was only marginally
significant, as shown in Table 5. The multiple imputation analyses
paediatric
obtained similar results, but with a statistically significant associ‐
General

0.091

0.359
ation between job insecurity and job strain (Appendix S1). Results
ICU

16
from the model with job strain as a continuous variable were consis‐
tent (Appendix S2).
intermediate
Paediatric

0.000

0.188
ICU

4 | D I S CU S S I O N
cardiovascular

The results of this study show a high percentage of nurses with

job strain (more than half) in comparison with other studies (Lee &
0.000

0.188
Adult

High
ICU

Wang, 2002; Lu et al., 2007). The findings also demonstrate that job
8

strain among nurses (nursing assistants and nursing professionals)

hospitalization

is significantly associated with months employed, educational level,

Paediatric

age, social support, and high use of patient assessment scales.

0.000

0.183

The high percentage of job strain found in this study could be

7

explained by conditions in the institution where this investigation

cardiovascular

was conducted. For example, this institution provides highly complex

care and nearly all of its units are ICUs, which demand more attention
Paediatric

and care activities from nurses. These units have also been shown
0.013

0.183
ICU

to have a higher risk of job strain than other hospital units (Zhang
6

et al., 2013). Other conditions in Colombia could also provide some

Transplant

insight into job strain among nurses, such as a complex labour market
0.000

0.180

for health professionals characterized by indirect engagement with

ICU

imputed to patient assessment scales that were not mentioned by nurses in the respective hospital unit

health agencies, wage gaps between nurses and other health profes‐
Intermediate

sionals, longer work hours, unemployment, limited use of skills, and

higher psychological demands (Acosta, 2009; Romero, 2007).
0.000

0.164
Adult

The socio‐demographic characteristics associated with job strain

ICU

in this investigation were similar to other studies (Urbanetto et al.,

Moderate

2011). Seniority could be considered an opportunity to increase

Urgency

0.000

0.156

experience, but this study found that working more months at the
6

institution was related to prolonged exposure to stressors that con‐

tributed to job strain.
hospitalization

In relation to nurses’ educational level, this study showed that

specialized nurses (nursing professionals) who were likely to carry
0.000

0.153
Adult

out administrative tasks had a lower risk of job strain than technical
6

nurses (nursing assistants) who likely had less control and were in
Neonatal

charge of the direct care of patients and their families. This finding
0.000

0.125

suggests the need to improve the work environment, especially for

Surgery ICU

nursing assistants.
It was not possible to draw any conclusions about gender dif‐
0.000

0.000

ferences among the participants in the study. This is a reflection of

0

nursing in Colombia, where it is uncommon to find men in this role.

Ambulatory

Further studies are needed to explore the relationship between the

0.000

0.000

role of gender and job strain in Colombia and Latin America.

Low
unit

ICU: intensive care unit.

0
TABLE 2 (Continued)

The association between social support and job strain is consis‐

Relevance of use

tent with the results described previously by other investigations

Nursing scales

Braden Scale

Mean Smith's

conducted with nursing staff worldwide (Urbanetto et al., 2011).

of patient's
evaluation
Scales (n)

salience

The inverse association found between job strain and receiv‐

Pediatric

scales

ing social support from the head of nursing and from colleagues
reflects the positive role of the workgroup in fostering teamwork
 |
48       TRIANA‐PALENCIA et al.

TA B L E 3   Nursing workforce and

Worker
turnover of nursing staff, by hospital unit
resignations, n Turnover of
Hospital units Total workers (%) P value nursing staff

Neonatal ICU 168 26 (15.48) Reference High

General paediatric ICU 68 9 (13.24) 0.661
Ambulatory service 117 13 (11.11) 0.292
Adult cardiovascular ICU 119 12 (10.08) 0.184
Transplant ICU 113 11 (9.73) 0.163
Adult hospitalization 199 16 (8.04) 0.026 Low
Pediatric cardiovascular ICU 266 21 (7.89) 0.013
Surgery 97 7 (7.22) 0.050
Adult intermediate ICU 91 6 (6.59) 0.038
Paediatric intermediate ICU 56 3 (5.36) 0.065
Paediatric hospitalization 108 3 (2.78) 0.001
Urgency 53 1 (1.89) 0.007

Cardiovascular Foundation of Colombia, 2009 to 2012 (n = 1,455).

P value: statistical significance of the chi‐squared test or the Fisher’s exact test with neonatal ICU as
reference (highest turnover of nursing staff).
ICU: intensive care unit.

and collaboration, knowledge acquisition, and the wellbeing of col‐ cardiovascular ICU. This may be related to the pressure experi‐
leagues. Although the literature related to the job strain model usually enced by unit coordinators to collect all of the information needed
does not indicate that high psychological demands and low control to provide quality indicators. The present study found that hos‐
are direct consequences of a lack of social support, systems theory pital units with a high use of patient assessment scales are units
provides the opportunity to consider the effects of some elements with high psychosocial risk. Hospital units with a high use of these
when others in the same system change. The strengthening of social scales also reported high turnover of nursing staff, as was the case
support and interpersonal relationships among nursing staff in health in the general paediatric ICU and adult cardiovascular ICU. This
institutions could be an effective strategy to mitigate job strain. This suggests a possible association between psychological demands
is particularly relevant given the subordinate relationships among and turnover of nursing staff, a hypothesis to be tested in future
coordinator‐nursing head‐nursing assistants, which could create in‐ studies in Colombia.
terpersonal conflicts (Urbanetto et al., 2011) and possibly affect not The present study did not find an association between job strain
only nurses but also the quality of service and the institution’s image. and turnover of nursing staff. This difference between our results
Strengthening social support among nurses could also mitigate the and the available evidence may have occurred because the present
effect of job strain on health, as has been mentioned in the literature study evaluated nursing staff turnover as a contextual effect while
(Garcia‐Rojas, Choi, & Krause, 2015; Schmidt, 2013). the other studies evaluated it at the individual level (Hayes et al.,
Although patient assessment scales are useful for nursing pro‐ 2012). In addition, turnover of nursing staff could be an orthogo‐
cesses during the evaluation stage, the present study’s finding of an nal factor in job strain, directly contributing to health problems in‐
association between their high use and job strain suggests the need dependently of job strain. Contextual conditions in the workplace
to review whether there are enough nurses in each hospital unit to should be further assessed in Colombia to approximate the effect of
perform this task. If the staff is not sufficient, it would be important working conditions on the health and wellbeing of nurses.
to evaluate whether shift personnel are experiencing work overloads, Some of the strengths of this study are its exploration of the
as this could adversely affect the quality of the services provided way in which job strain was associated with socio‐demographic
by the institution, including poor communication between patients characteristics and the relationship between job strain and certain
and nurses, less time to adequately evaluate the care provided, and work conditions characteristic of the current global market, such
increased hospital stays (Elliott, Young, Brice, Aguiar, & Kolm, 2014). as turnover among Colombian nurses. The analysis also included
The association between high use of patient assessment scales and subjective (self‐reports) and objective indicators (e.g., hospital
job strain also highlights the need to monitor and evaluate the use of records).
scales in the institution to verify whether they are being used correctly This study also had some limitations. First, temporal criteria
or treated as just one more quality‐control requirement to be met. could not be included due to the study’s cross‐sectional design and,
The results indicate a high use of patient assessment scales in therefore, the possibility of reverse causality cannot be discarded.
the general paediatric ICU, paediatric intermediate ICU, and adult Nonetheless, as a first approach to the subject, this study contributes
TRIANA‐PALENCIA et al. |
      49

TA B L E 4   Crude logistic regression model with robust variance TA B L E 5   Adjusted logistic regression model with robust
estimator: job strain and sociodemographic characteristics, social variance estimator: job strain and socio‐demographic
support, job insecurity, use of patient assessment scales and characteristics, social support, job insecurity, use of patient
turnover of nursing staff assessment scales

Job strain, Crude OR Job strain,

Characteristic (95% CI) P value Adjusted OR (95%
Characteristic CI) P value
Individual level
Gender Individual level

Female Reference Age

Male 0.56 (0.21–1.53) 0.261 ≤29 years Reference

Position 30–35 years 0.63 (0.38–1.07) 0.087

Head of nursing or Reference ≥36 years 0.37 (0.15–0.91) 0.029

coordinator Educational level
Nursing assistant 1.43 (0.77–2.68) 0.257 Technical (nursing assistant) Reference
Marital status Professional (nursing 0.86 (0.46–1.59) 0.620
Single Reference professional)

Married 0.89 (0.42–1.89) 0.759 Specialized (nursing 0.36 (0.18–0.75) 0.006

professional)
Other 1.33 (0.66–2.71) 0.426
Months employed
Age
Less than 60 months Reference
≤29 years Reference
More than 60 months 2.97 (1.74–5.06) 0.000
30–35 years 0.75 (0.46–1.23) 0.256
Social support 0.88 (0.80–0.98) 0.019
≥36 years 0.72 (0.36–1.42) 0.342
Job insecurity 1.26 (0.97–1.63) 0.090
Educational level
Hospital unit
Technical (nursing Reference
assistant) Use of patient assessment
scales
Professional (nursing 0.99 (0.52–1.89) 0.985
professional) Low Reference

Specialized (nursing 0.44 (0.23–0.84) 0.013 Moderate 1.82 (0.75–4.44) 0.188

professional) High 2.62 (1.34–5.13) 0.005
Months employed Cardiovascular Foundation of Colombia, 2013 (n = 222).
Less than 60 months Reference
More than 60 months 1.59 (0.94–2.69) 0.083 risk of selection bias, given the lack of participation by 35% of the
Number of children 1.03 (0.815–1.31) 0.782 nurses in the institution. Although vacations, medical leave, and lack
Social support 0.88 (0.79–0.98) 0.024 of interest were identified as the main reasons for non‐participation,

Job insecurity 1.28 (1.02– 1.60) 0.035 the association between these reasons and job strain is uncertain,
especially with regard to the latter two. It is worth noting that some
Hospital unit
of the hospital units that had a low response rate also reported a
Use of patient assess‐
ment scales high use of patient assessment scales, including the general paedi‐

Low Reference
atric ICU and the paediatric intermediate ICU. Assuming a higher
probability of job strain among nurses who did not participate in
Moderate 1.94 (0.88–4.27) 0.099
the study, the association between a high use of patient assessment
High 2.27 (1.10–4.70) 0.026
scales and job strain may therefore be stronger. Another limitation
Turnover of nursing staff 0.96 (0.88–1.04) 0.302
of this study was the number of informants who responded to free
Cardiovascular Foundation of Colombia, 2013 (n = 222). lists. The literature suggests that the sample size for free lists de‐
pends on the coherence of the domain, such that only one informant
to understanding how current quality standardization processes in would be necessary if there were perfect agreement among the re‐
hospitals may affect nursing staff. The second limitation was the use spondents (Gravlee, 2002; Sinha, 2003). As the number of key nurse
of self‐reports to evaluate job strain, job insecurity, social support, informants was not chosen according to the degree of coherence,
and the use of patient assessment scales. Nevertheless, the JCQ the number of key nurse informants participating in the free lists in
used in this study has been validated for the Colombian population some hospital units may be less than required. Despite this, the free
and was shown to be a suitable instrument to assess psychosocial list exercise used in this study could be an option for exploring oc‐
risk in this country (Gómez‐Ortíz, 2011). A third limitation was the cupational or environmental factors in future studies. Finally, other
 |
50       TRIANA‐PALENCIA et al.

characteristics of individuals and hospital units, such as frequency Cedillo, L., & Karasek, R. (2003). Reliability and Validity of the Spanish
of assignment to shifts, whether nurses had other jobs, leisure time Version of the Job Content Questionnaire among Maquiladora Women
Workers. Retrieved from www.jcqcenter.org
physical activity, and the patient/nurse ratio, were not assessed in
Chen, S.‐Y., Wu, W.‐C., Chang, C.‐S., & Lin, C.‐T. (2015). Job rotation
this study and should be explored in future investigations. and internal marketing for increased job satisfaction and organi‐
sational commitment in hospital nursing staff. Journal of Nursing
Management, 23(3), 297–306. https://doi.org/10.1111/jonm.12126
Chiu, Y.‐L., Chung, R.‐G., Wu, C.‐S., & Ho, C.‐H. (2009). The effects of
5 |  I M PLI C ATI O N S FO R N U R S I N G
job demands, control, and social support on hospital clinical nurses’
M A N AG E M E NT intention to turn over. Applied Nursing Research, 22(4), 258–263.
https://doi.org/10.1016/j.apnr.2008.02.006
The findings of this study suggest possible opportunities for managers da Silva, J. L. L., Soares, R. d. S., Costa, F. D. S., Ramos, … L. R.
(2015). Psychosocial factors and prevalence of burnout syn‐
to improve nursing processes, the work conditions of nursing staff,
drome among nursing workers in intensive care units. Revista
and the quality of institutions. Further studies need to be carried out Brasileira de Terapia Intensiva, 27(2), 125–133. https://doi.
to confirm and to understand better the associations among the stud‐ org/10.5935/0103-507X.20150023
ied variables. Clear guidelines that establish the size of nursing teams Dewa, C. S., Loong, D., Bonato, S., Thanh, N. X., & Jacobs, P. (2014).
How does burnout affect physician productivity? A systematic lit‐
according to the number and complexity of patients are necessary in
erature review. BMC Health Services Research, 14(325). https://doi.
the institution and Colombia. Managers should also consider imple‐
org/10.1186/1472-6963-14-325
menting programmes to increase the autonomy of nurses, their skill Echeverri, E., & Vélez, D. (2005). La enfermería, una profesión de servi‐
use and their social support. Strategies that have been successful in cio. Revista Actualizaciones en Enfermería, 8(4), 19–21.
other countries and which could be implemented include internal mar‐ Elliott, D. J., Young, R. S., Brice, J., Aguiar, R., & Kolm, P. (2014). Effect
of hospitalist workload on the quality and efficiency of care.
keting (Chen, Wu, Chang, & Lin, 2015), decreasing the patient/nurse
JAMA Internal Medicine, 174(5), 786–793. https://doi.org/10.1001/
ratio (Needleman, Buerhaus, Mattke, Stewart, & Zelevinsky, 2002), jamainternmed.2014.300
and increasing the number of nurses with higher education levels Fernandes Portela, L., Luna, C. K., Rotenberg, L., … R. H. (2015). Job
(Aiken, Clarke, Sloane, Lake, & Cheney, 2009). strain and self‐reported insomnia symptoms among nurses:
What about the influence of emotional demands and social sup‐
port? BioMed Research International, 2015(820610). https://doi.
org/10.1155/2015/820610
AC K N OW L E D G E M E N T S
Fiabane, E., Giorgi, I., Sguazzin, C., & Argentero, P. (2013). Work en‐
The authors would like to thank the managers and nurses at the gagement and occupational stress in nurses and other healthcare
workers: The role of organisational and personal factors. Journal
Cardiovascular Foundation of Colombia for their collaboration in
of Clinical Nursing, 22(17–18), 2614–2624. https://doi.org/10.1111/
this study. jocn.12084
Garcia‐Rojas, I. J., Choi, B., & Krause, N. (2015). Psychosocial job fac‐
tors and biological cardiovascular risk factors in Mexican workers.
E T H I C A L A P P R OVA L American Journal of Industrial Medicine, 58(3), 331–351. https://doi.
org/10.1002/ajim.22410
Verbal informed consent was obtained from the participants. This Gómez‐Ortíz, V. (2011). Assessment of psychosocial stressors at work:
study was approved by the ethics committee of the Cardiovascular Psychometric properties of the JCQ in Colombian workers. Revista
Foundation of Colombia with the record number 450. Latinoamericana de Psicología, 43(2), 329–342.
Gravlee, L. (2002). The uses and limitations of free listing in ethnographic re‐
search. Retrieved from https://gravlee.org/ang6930/freelists.htm.
Hayes, L. J., O’Brien‐Pallas, L., Duffield, C., … N. (2012). Nurse turnover: A
REFERENCES
literature review—an update. International Journal of Nursing Studies,
Acosta, J. H. (2009). Labor conditions for health professionals starting 49(7), 887–905. https://doi.org/10.1016/j.ijnurstu.2011.10.001
from the Act 100 of 1993: Evolution and a case study for Medellin. Henriques da Silva, F. (2015). Estresse no trabalho de enfermeiros que
Revista Gerencia y Políticas de Salud, 8(16), 107–131. atuam em hospitais privados segundo o modelo demanda‐controle.
Aiken, L. H., Clarke, S. P., Sloane, D. M., Lake, E. T., & Cheney, T. (2009). Retrieved from https://bases.bireme.br/cgi-bin/wxislind.exe/iah/
Effects of hospital care environment on patient mortality and nurse online/?IsisScript=iah/iah.xis&src=google&base=LILACS&xm‐
outcomes. Journal of Nursing Administration, 39(7–8 Suppl), S45– l:lang=p&nextAction=lnk&exprSearch=758261&indexSearch=ID.
S51. https://doi.org/10.1097/NNA.0b013e3181aeb4cf Jerome‐Smith, J. (1993). Using ANTHOPAC 3.5 and a spreadsheet to
Araújo, T. M., Aquino, E., Menezes, G., Santos, C. O., & Aguiar, L. (2003). compute a free‐list salience index. Cultural Anthropology Methods,
Aspectos psicossociais do trabalho e distúrbios psíquicos entre 5(3), 1–3.
trabalhadoras de enfermagem. Revista de Saúde Pública, 37(4), 424– Juárez‐García, A. (2007). Psychosocial work factors associated to blood
433. https://doi.org/10.1590/S0034-89102003000400006 pressure and cardiovascular symptoms among Mexican nurses.
Bosi de Souza Magnago, T. S., Luz Lisboa, M. T., Griep, R. H., Zeitoune, Salud Publica de México, 49(2), 109–117.
… J. P. 2010). Working conditions of nurses: Evaluation based on Juárez‐García, A., Aldo, V. C., Viviola, G. O., Carlo, C., & Peter, S. (2008).
the demand‐ control model. Acta Paulista de Enfermagem, 23(6), El Modelo Demanda/Control y la Salud Mental en Profesionales de la
811–817. Salud: Un Estudio en Tres Países Latinoamericanos, in Segundo foro de
Buchan, J. (2010). Reviewing the benefits of health workforce las Américas en investigación sobre factores psicosociales. Retrieved
stability. Human Resources for Health, 8(29), https://doi. from https://www.factorespsicosociales.com/segundoforo/simpo‐
org/10.1186/1478-4491-8-29 sio/Juarez-Vera-Gomez-etal.pdf.
TRIANA‐PALENCIA et al. |
      51

Juárez‐García, A., Calzaretta, A. R. V., Soto, C. M., Ortíz, V. G., Feldman, L., emergency, anesthesiology and intensive care units depends on
& Mendoza, E. H. (2014). Demanda/control y la salud mental en pro‐ qualification: A Job Demand‐Control survey. International Archives
fesionales de la salud: Un estudio en seis países Latinoamericanos. of Occupational and Environmental Health, 89(2), 221–229. https://
Informacio Psicológica, 2(108), 2–18. doi.org/10.1007/s00420-015-1065-7
Karasek, R. (1979). Job demands, job decision latitude, and mental strain: Trybou, J., Germonpre, S., Janssens, H., Casini, A., Braeckman, L., De
Implications for job redesign. Administrative Science Quarterly, 24(2), Bacquer, D., & Clays, E. (2014). Job‐related stress and sickness
285–308. https://doi.org/10.2307/2392498 absence among belgian nurses: A prospective study. Journal of
Kivimäki, M., Nyberg, S. T., Batty, G. D., Fransson, E. I., Heikkilä, K., Nursing Scholarship, 46(4), 292–301. https://doi.org/10.1111/
Alfredsson, L. … IPD‐Work Consortium (2012). Job strain as a risk jnu.12075
factor for coronary heart disease: A collaborative meta‐analysis of Tsai, Y.‐C., & Liu, C.‐H. (2012). Factors and symptoms associated with
individual participant data. The Lancet, 380(9852), 1491–1497. work stress and health‐promoting lifestyles among hospital staff:
Lee, I., & Wang, H.‐H. (2002). Perceived occupational stress and related A pilot study in Taiwan. BMC Health Services Research, 12(199).
factors in public health nurses. Journal of Nursing Research, 10(4), https://doi.org/10.1186/1472-6963-12-199
253–260. https://doi.org/10.1097/01.JNR.0000347606.91295.76 Twigg, D., & McCullough, K. (2014). Nurse retention: A review of strate‐
Leguizamón, L. C., & Gómez Ortíz, V. (2002). Condiciones laborales y de gies to create and enhance positive practice environments in clin‐
salud en enfermeras de Santafé de Bogotá. International Journal of ical settings. International Journal of Nursing Studies, 51(1), 85–92.
Clinical and Health Psychology, 2(1), 173–182. https://doi.org/10.1016/j.ijnurstu.2013.05.015
Lu, H., While, A. E., & Barriball, K. L. (2007). Job satisfaction and its re‐ Urbanetto, J. d. S., Silva, P. C., Hoffmeister, E., de Negri, B. S., … C. E.
lated factors: A questionnaire survey of hospital nurses in main‐ (2011). Workplace stress in nursing workers from an emergency
land China. International Journal of Nursing Studies, 44(4), 574–588. hospital: Job Stress Scale analysis. Revista Latino‐Americana
https://doi.org/10.1016/j.ijnurstu.2006.07.007 de Enfermagem, 19(5), 1122–1131. https://doi.org/10.1590/
Needleman, J., Buerhaus, P., Mattke, S., Stewart, M., & Zelevinsky, K. S0104-11692011000500009
(2002). Nurse‐staffing levels and the quality of care in hospitals. Urbanetto, J. d. S., Silva, P. C., Hoffmeister, E., de Negri, B. S., … C. E.
New England Journal of Medicine, 346(22), 1715–1722. https://doi. (2013). Work‐related stress according to the demand‐control model
org/10.1056/NEJMsa012247 and minor psychic disorder in nursing workers. Revista da Escola de
Pinho, P. d. S., & Araújo, T. M. d. (2007). Trabalho de enfermagem em uma Enfermagem da USP, 47(5), 1186–1193.
unidade de emergência hospitalar e transtornos mentais. Revista Wu, H., Chi, T.‐S., Chen, L. i., Wang, L., & Jin, Y.‐P. (2010).
Enfermagem UERJ, 15(3), 329–336. Occupational stress among hospital nurses: Cross‐sectional sur‐
Romero, B. M. N. (2007). El cuidado en la adversidad: Compromisos de re‐ vey. Journal of Advanced Nursing, 66(3), 627–634. https://doi.
sistencia desde la docencia. Avances en Enfermería, 25(2), 112–123. org/10.1111/j.1365-2648.2009.05203.x
Schmidt, D. R. C. (2013). Demand‐Control model and occupational stress Zhang, A., Tao, H., Ellenbecker, C. H., & Liu, X. (2013). Job satisfaction in
among nursing professionals: Integrative review. Revista Brasileira mainland China: Comparing critical care nurses and general ward
de Enfermagem, 66(5), 779–788. nurses. Journal of Advanced Nursing, 69(8), 1725–1736. https://doi.
Sinha, R. (2003). Beyond cardsorting: Free‐listing methods to explore user org/10.1111/jan.12033
mental models, Boxes and Arrows. Retrieved from https://boxesan‐
darrows.com/beyond-cardsorting-free-listing-methods-to-ex‐
plore-user-categorizations/. S U P P O R T I N G I N FO R M AT I O N
Stansfeld, S., & Candy, B. (2006). Psychosocial work environment and
mental health–a meta‐analytic review. Scandinavian Journal of Work, Additional supporting information may be found online in the
Environment and Health, 32(6), 443–462. https://doi.org/10.5271/ Supporting Information section at the end of the article. 
sjweh.1050
Tavares, J. P., Beck, C. L. C., Magnago, T. S. B. S., Zanini, R. R., & … L.
(2012). Minor psychiatric disorders among nurses university fac‐
ulties. Revista Latino‐Americana de Enfermagem, 20(1), 175–182. How to cite this article: Triana‐Palencia E, Cárdenas‐
https://doi.org/10.1590/S0104-11692012000100023 Cárdenas LM, Juárez‐García A, Quiroz‐Muysina J, Idrovo AJ.
Theme Filha, M. M., De Souza Costa, M. A., & Rodrigues Guilam, M. C.
Use of assessment scales, turnover and job strain in nursing
(2013). Estrés ocupacional y autoevaluación de salud entre profesio‐
nales de enfermería. Revista Latino‐Americana de Enfermagem, 21(2).
staff: A study in a Colombian hospital. J Nurs Manag.
Trousselard, M., Dutheil, F., Naughton, G., Cosserant, S., Amadon, S., 2019;27:42–51. https://doi.org/10.1111/jonm.12647
Dualé, C., & Schoeffler, P. (2016). Stress among nurses working in