Content uploaded by Mark Hancock
Author content
All content in this area was uploaded by Mark Hancock on Jun 30, 2019
Content may be subject to copyright.
Research
Recurrence of low back pain is common: a prospective inception cohort study
Tatiane da Silva
a
, Kathryn Mills
a
, Benjamin T Brown
b
, Natasha Pocovi
a
, Tarcisio de Campos
a
,
Christopher Maher
c
, Mark J Hancock
a
a
Department of Health Professions, Macquarie University, Sydney;
b
Department of Chiropractic, Macquarie University, Sydney;
c
Institute for Musculoskeletal Heath, Sydney
School of Public Health, The University of Sydney, Australia
KEY WORDS
Risk
Recurrence
Low back pain
Prognosis
Cohort studies
ABSTRACT
Questions: How commonly and how quickly does low back pain reoccur in a cohort of people who have
recently recovered from an episode of low back pain? What are the prognostic factors for a recurrence of low
back pain? Design: Prospective inception cohort study with monthly follow-up for 12 months. Participants:
A total of 250 patients who had recovered from an episode of low back painwithin the last month. Outcome
measures: The primary outcome was days to recurrence of an episode of low back pain. Secondary outcomes
were: days to recurrence of low back pain severe enough to limit activity moderately, and days to recurrence
of low back pain for which healthcare was sought. Results: Within 12 months after recovery, 69% (95% CI 62
to 74) of participants had a recurrence of an episode of low back pain, 40% (95% CI 33 to 46) had a recurrence
of activity-limiting low back pain, and 41% (95% CI 34 to 46) had a recurrence of low back pain for which
healthcare was sought. The median time to recurrence of an episode of low back pain was 139 days (95% CI
105 to 173). Frequent exposure to awkward postures, longer time sitting (.5 hours per day), and more than
two previous episodes were predictive of recurrence of an episode of low back pain within 12 months (p,
0.01). Conclusion: Recurrence of low back pain is very common, with more than two-thirds of individuals
having a recurrence within 12 months after recovery. Prognostic factors for a recurrence include exposure to
awkward posture, longer time sitting, and more than two previous episodes. [da Silva T, Mills K, Brown BT,
Pocovi N, de Campos T, Maher C, Hancock MJ (2019) Recurrence of low back pain is common: a pro-
spective inception cohort study. Journal of Physiotherapy -:-–-]
© 2019 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under
the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Introduction
The majority of patients presenting with acute low back pain
(LBP) recover quickly;
1
however, recurrences are believed to be
common and are likely to be responsible for much of the burden
associated with LBP.
2,3
A recent systematic review investigating the
risk of a recurrence of LBP in patients who have recovered from a
previous episode concluded that it is not yet possible to obtain robust
estimates of the risk of recurrence, due to the heterogeneity and
generally poor methodological quality of the small number of exist-
ing studies.
4
The review also found very little evidence about prog-
nostic factors for recurrence of LBP. The presence of previous
episodes of LBP was the only consistent prognostic factor for
recurrence.
5,6
Most previous studies
7–10
investigating recurrence of LBP
included survival cohorts of people who recovered from a previous
episode of LBP at different times in the past, and may have been
recovered for long and variable periods. Studies based on survival
cohorts may produce biased estimates because the prognosis of
people who have had a condition for a long time is likely to be
different from those who recently developed the condition.
11
Inception cohort studies enrol participants at an early and uniform
time point in the course of the condition and follow them to observe
if an event of interest occurs (eg, death or recurrence).
11,12
Inception
cohort studies therefore avoid the bias associated with survival
cohorts.
11
Another limitation of previous studies is the use of different
recurrence definitions or the lack of any clear definition of a
recurrence.
5–9
Different definitions are likely to produce different
estimates of recurrence. A recent consensus document defined a
recurrence of an episode of LBP as ‘a return of LBP lasting at least 24
hours with a pain intensity of .2 on an 11-point numerical rating
scale following a period of at least 30 days pain-free’.
13
Furthermore,
most previous studies
7–10,14,15
do not clearly describe how recovery
from the previous episode was defined, so it is unclear whether
participants had recovered and were therefore truly at risk of
recurrence, or for how long they had recovered prior to entering the
study. Therefore, a large, well-designed inception cohort study to
provide reliable estimates of the risk of recurrence and to identify
prognostic factors is needed.
Therefore, the research questions for this inception cohort study
were:
1. How commonly and how quickly does LBP reoccur in a cohort of
people who have recently recovered from an episode of LBP?
2. What are the prognostic factors for a recurrence of LBP?
Journal of Physiotherapy -(2019) -–-
https://doi.org/10.1016/j.jphys.2019.04.010
1836-9553/© 2019 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/
licenses/by-nc-nd/4.0/).
journal homepage: www.elsevier.com/locate/jphys
Methods
Design
This prospective inception cohort study recruited patients who
had recovered from an episode of LBP within the past month and who
had been discharged from primary care practices in Sydney, Australia,
from August 2015 to August 2017. Physiotherapists and chiropractors
practising within metropolitan Sydney screened consecutive patients
with LBP for eligibility. Patients were informed about the study and
invited to either contact the researchers directly or provide their
contact details if they were interested in finding out more about the
study. Potential participants were then contacted by telephone to
discuss the details of the study and assess their eligibility for
participation. Participants were followed monthly for 12 months.
Participants
Inclusion criteria were age 18 years, and recovery within the last
month from a previous episode of non-specific LBP with or without
associated leg pain or radiculopathy. Non-specific LBP was defined as
pain in the area between the 12th rib and buttock crease not
attributed to a specific diagnosis (eg, ankylosing spondylitis, vertebral
fracture).
16–18
Recovery was defined as a score of 0 or 1 on an 11-point
numerical rating scale for 7 consecutive days.
19,20
If individuals were
willing to participate when initially contacted but were not yet
recovered, they were followed up fortnightly until they were
recovered. Individuals were excluded if they met any of the following
criteria: previous spinal surgery, spinal surgery scheduled in the
following 12 months, or inadequate English comprehension to
complete the outcome measures.
Baseline measures
Participants had the option of completing the baseline assessment
either by telephone or as an online Qualtrics survey.
21
At baseline,
demographic data and a range of potential prognostic factors for
recurrence of LBP were collected. Fifteen a priori prognostic factors
were chosen based on the existing literature or biological plausibility
of a factor being predictive of a recurrence of LBP.
4,22
Table 1 describes
all prognostic factors that were investigated, how they were
measured, and how they were coded in the analyses. The prognostic
factors were measured using previously published
19,22–25
and
validated questionnaires.
26,27
Outcome measures
The primary outcome was days to recurrence of an episode of LBP.
Recurrence of an episode of LBP was defined according to the
consensus definition as ‘return of LBP lasting at least 24 hours with a
pain intensity of .2 on an 11-point numerical rating scale’.
13
Secondary outcomes were: days to recurrence of LBP severe enough
to limit activity, and days to recurrence of LBP for which healthcare
was sought. Recurrence of activity-limiting LBP was defined as a
recurrence of an episode of LBP causing moderate or greater activity
limitation measured using an adaptation of Item 8 of the 36-Item
Short Form Survey (SF-36).
27
This was measured using the
question: ‘During the recurrence, how much did low back pain
interfere with your normal work (including work both outside the
home and housework)?’The response options were: not at all, a little
bit, moderately, quite a bit, and extremely. Recurrence of LBP causing
care-seeking was defined as recurrence of an episode of LBP resulting
in a consultation with a healthcare provider.
Follow-up
Participants were contacted monthly by email or text message
(based on the participant’s preference) for 12 months. Participants
were asked if they had a recurrence of LBP lasting at least 24 hours
and with a pain intensity of .2 on an 11-point numerical rating scale.
If a participant reported a recurrence, they were contacted by
telephone to obtain a detailed description of the episode. Participants
not responding to monthly follow-up within 48 hours were contacted
by telephone. If a recurrence was reported, participants were no
longer followed monthly, but continued to complete follow-up
assesments at 3, 6, 9 and 12 months. Using this approach, the
Table 1
Candidate prognostic factors for a recurrence of LBP.
Candidate prognostic factor Data format at collection Coding of data in the model
Age Age in years (continuous) Continuous
Body mass index Calculated as height in m divided by the square of weight in kg (continuous) Continuous
Smoking history How would you describe your cigarette smoking? (never, used to smoke but
have quit, or current smoker)
Categorical: Never; Used to smoke but have quit;
Current smoker
Exposure to heavy loads How often are you engaged in any manual task involving heavy loads?
(never; very rarely, rarely, occasionally, frequently, very frequently)
Categorical: Rarely (rarely, very rarely or never);
Occasionally; Frequently (frequently or very frequently)
Exposure to awkward posture How often are you engaged in any manual task or activity involving an
awkward position? (never, very rarely, rarely, occasionally, frequently, or
very frequently)
Categorical: Rarely (rarely, very rarely or never);
Occasionally; Frequently (frequently or very frequently)
Physical activity The Active Australia Questionnaire (vigorous, moderate, or low) Categorical: Vigorous; Moderate; and Low
Time sitting Could you indicate how many hours you spend sitting on an average
weekday, including sitting for travel, work and leisure? (continuous)
Categorical: 0 to 5 hours; .5 hours
General health In general, would you say your health is: (excellent, very good, good,
fair, or poor)
Categorical: Excellent (excellent or very good);
Good; and Poor (fair or poor)
Number of previous episodes How many previous episodes of low back pain have you had? (continuous) Categorical: 1 to 2 episodes; 3 to 10 episodes; .10 episodes
Duration of last episode How long did your most recent episode of low back pain last (days)?
(continuous)
Categorical: ,2 weeks; 2 to 6 weeks; 6 to 12 weeks
Perceived risk of recurrence In your view, how large is the risk that you will develop a recurrence
of low back pain in the following 12 months? Considering a 0-to-10
numerical scale. (Continuous)
Categorical: 0 to 5 points; .5 points
Depression DASS21 (normal, mild, moderate, severe, or extremely severe) Categorical: Normal (normal or mild); Moderate
(moderate, severe or extremely severe)
Anxiety DASS21 (normal, mild, moderate, severe, or extremely severe) Categorical: Normal (normal or mild); Moderate
(moderate, severe or extremely severe)
Stress DASS21 (normal, mild, moderate, severe, or extremely severe) Categorical: Normal (normal or mild); Moderate
(moderate, severe or extremely severe)
Sleep quality During the past week, how would you rate your sleep quality overall?
(very good, fairly good, fairly bad, or very bad)
Categorical: Good (very good or fairly good); Bad
(fairly bad or very bad)
DASS-21 = Depression, Anxiety and Stress Scale - 21 Items, LBP = low back pain.
2da Silva et al: Risk of recurrence of low back pain
number of days to first recurrence was able to be determined for each
of the three recurrence definitions.
Data analysis
Sample size calculation was based on the two aims related to
recurrence proportions and prognostic factors for a recurrence. For
the calculation related to recurrence proportion, considering the
recurrence estimate of 33% based on previous studies,
6,28
and a
required precision of the sample estimate within 6 absolute
percentage points, that is a 95% confidence interval from 27 to 39%, a
sample size of 236 was required. With respect to the aim of
investigating prognostic factors, previous studies suggest at least 10
events per candidate variable in the multivariate model.
29,30
Based
on the recurrence estimate of 33%, a sample of 250 participants
would result in 83 events enabling eight candidate predictor
variables to be investigated in the model. Therefore, eight predictor
variables were pre-specified from the initial list of 15: age, exposure
to heavy loads, exposure to awkward posture, physical activity, time
sitting, number of previous episodes of LBP, perceived risk of
recurrence of LBP, and depression. However, if the recurrence
proportion was higher, it was planned to investigate up to 15
variables, depending on the number of events. Therefore, a sample
size of 250 was used.
We censored participants for whom follow-up data were
missing or who did not have a recurrence at the time of their last
follow-up. Censoring allows use of data from people who do not
experience the outcome during the time period that they were
followed.
31,32
Survival curves plotting days to recurrence were used to describe
the proportion of people who had a recurrence, considering each of
the three definitions over 1 year. Time to recurrence was treated in
two ways: ‘inception time uncorrected’, which considered the time of
study entry as the reference time; and ‘inception time corrected’,
which transformed the time data by adding the number of days that
patients were recovered before entering the study, as this was
available for all participants.
1
Inception time uncorrected was
considered the primary outcome.
Cox regression was used to investigate prognostic factors for
recurrence. First, some of the variables (exposure to heavy loads,
exposure to awkward posture, number of previous episodes,
depression, general health, duration of previous episodes, anxiety,
stress, and sleep quality) were recoded into ordinal categories
(Table 1). Visual inspection of survival curves and Cox regression
with time as the dependent variable were used to check the
proportional hazards assumption for each variable. Linearity of
continuous variables was assessed via visual inspection of survival
curves. We first ran univariate Cox regression models to test for an
association between each individual variable and time to recurrence.
These were used to understand univariate associations and help
interpretation of the multivariate model eligibility for the
multivariate model.
33
Multivariate Cox regression analysis was then
conducted using a backward selection procedure with p-values of
,0.05 to enter the model and .0.10 to exit the model.
Completeness of follow-up was calculated using the completeness
index.
31
This index indicates the sum of follow-up times divided by
the sum of potential follow-up times. A completeness index value of
100% indicates complete follow-up. All analyses were performed
with commercial software.
Results
Flow of participants through the study
A total of 409 consecutive potential participants were referred
from 22 primary care practices (19 physiotherapists and nine
chiropractors). Of the 409 potential participants, 250 met the
inclusion criteria and entered the study (Figure 1). In total, 236
participants (94%) were successfully followed until a recurrence or
were censored at the 12-month follow-up. The remaining 14 patients
were lost to follow-up during the 12-month follow-up period and
were censored early. Completeness of follow-up, according to the
completeness index, was 96% of person time.
Characteristics of participants
The mean age of participants was 50 years (SD 15); 50% were
male, and 79% were referred from a physiotherapist. The median
days from the date of recovery to the date of study entry was 14 days
(IQR 7 to 27.5). The median number of previous episodes was five
episodes (IQR 2 to 18.5), and the median duration for the previous
episode was 14 days (IQR 5 to 40.5). Table 2 presents baseline
demographic data for the study participants and Table 3 presents
their baseline data for the candidate prognostic factors, with
additional details for both sets of data provided in Table 4
(see eAddenda for Table 4).
Included and baseline data
collected (n = 250)
Screened for eligibility (n = 409)
Followed up until a recurrence or
censored at 12 months (n = 236)
Not included (n = 159)
•
•
•
•
•
•
•
•
•
•
•
declined to participate (n = 63)
uncontactable (n = 44)
had not recovered (n = 9)
chronic low back pain (n = 18)
pain elsewhere (n = 11)
spinal surgery (n = 5)
>1 month pain free (n = 4)
non-English speaker (n = 2)
pregnancy (n = 1)
moving overseas (n = 1)
age <18 years (n = 1)
Informed consent (n = 250)
Censored early (n = 14)
Figure 1. Flow of participants through the study.
Table 2
Baseline demographic characteristics of the study participants.
Variable Participants (n = 250)
Age (y), mean (SD) 50 (15)
Gender, n male (%) 125 (50)
Primary care clinician, n (%)
physiotherapist 198 (79)
Education level, n (%)
primary school 1 (,1)
some secondary school 1 (,1)
completed high school 56 (22)
some additional training 51 (20)
undergraduate university 84 (34)
postgraduate university 57 (23)
Research 3
Recurrence of an episode of low back pain
Using the uncorrected time, the cumulative probability of a
recurrence of an episode of LBP was 38% (95% CI 32 to 44) by 3
months, 56% (95% CI 49 to 62) by 6 months, and 69% (95% CI 62 to 74)
by 12 months. The median time to recurrence of an episode of LBP
was 139 days (95% CI 105 to 173) when calculated using the
uncorrected time, and 146 days (95% CI 114 to 178) using the
corrected time.
Recurrence of activity-limiting low back pain
Using the uncorrected time, the cumulative probability of a
recurrence of activity-limiting LBP was 17% (95% CI 13 to 22) by 3
months, 28% (95% CI 22 to 33) by 6 months, and 40% (95% CI 33 to 46)
by 12 months. Less than 50% of participants had a recurrence of
activity-limiting LBP by 1 year, so median time to recurrence could
not be calculated. The 75th percentile survival times (25% had
experienced a recurrence of activity-limiting LBP) was 152 days (95%
CI 104 to 201) when calculated using the uncorrected time and 163
days (95% CI 112 to 214) using the corrected time.
Recurrence of low back pain for which healthcare was sought
Using the uncorrected time, the cumulative probability of a
recurrence of LBP causing care-seeking was 12% (95% CI 8 to 17) by 3
months, 25% (95% CI 19 to 30) by 6 months, and 41% (95% CI 34 to 46)
by 12 months. Less than 50% of participants had a recurrence of LBP
causing care-seeking by 1 year, so median time to recurrence could
not be calculated. The estimate of 75th percentile survival times (25%
had experienced a recurrence) to a recurrence of LBP causing
care-seeking was 193 days (95% CI 139 to 247) when calculated using
the uncorrected time, and 207 days (95% CI 145 to 269) using the
corrected time.
Figure 2 presents the Kaplan-Meier survival curves for the
three definitions of recurrence, based on the uncorrected
inception time.
Prognostic factors for a recurrence of an episode of low back pain
As the recurrence of an episode of LBP proportion was higher than
expected (170 events), we were able to investigate 15 candidate
predictors (allowing at least 10 events per predictor). There was no
evidence that any of the continuous variables violated the linearity
assumption. However, the variables ‘time spent sitting’and
‘perceived risk of recurrence’violated the proportional hazards
assumption. Therefore, both variables were dichotomised using a
median split and the proportional hazards assumption was re-tested.
The dichotomised variables did not violate the proportional hazards
assumption. The results of the univariate analyses are shown in
Table 5. Of the 15 variables entered into the multivariate model,
exposure to awkward posture, time spend sitting, and number of
previous episodes were associated with recurrence of an episode of
LBP within 12 months. Participants who reported frequent exposure
to awkward posture had an 81% greater risk (HR 1.81, 95% CI 1.22 to
2.68) of having a recurrence than those who were rarely exposed to
awkward posture. However, there was no evidence that occasional
exposure to awkward posture increases the risk (HR 1.20, 95% CI 0.83
to 1.73) of having a recurrence. Participants who reported time spend
sitting .5 hours had 50% more risk (HR 1.50, 95% CI 1.08 to 2.09) of
having a recurrence than those who reported time spent sitting
between 0 and 5 hours. Compared with participants who reported
one or two previous episodes, participants who reported between
three and 10 previous episodes had 63% greater risk (HR 1.63, 95% CI
1.08 to 2.47), and those who reported more than ten previous
episodes had 94% greater risk (HR 1.94, 95% CI 1.28 to 2.94) of having
a recurrence. The results of the multivariate analyses are shown in
Table 6.
Discussion
The estimate of risk of recurrence of LBP in a representative
inception cohort from a primary care setting was much higher than
previously reported.
4,6,28
By 1 year, 69% of participants had a
recurrence of an episode of LBP, 40% of participants had a
recurrence of activity-limiting LBP, and 41% of participants had a
recurrence of LBP for which care was sought. It was also found that
exposure to awkward posture, a longer time spent sitting, and
more previous episodes of LBP were independent prognostic
factors associated with recurrence of an episode of LBP within 1
year.
This study overcomes many of the important limitations of
previous studies. It enrolled a large inception cohort of participants
who had recovered from a previous episode of LBP within the
previous month. Potential participants who still had low levels of
pain were followed until they were recovered, so they could be
enrolled soon after recovery. Monthly follow-ups were conducted
to avoid recall bias. This also allowed us to describe estimates of a
recurrence at different time points (eg, 3 or 6 months)
rather than just at 1 year. There was minimal loss to follow-up
Table 3
Baseline data of the study participants for the 15 candidate prognostic factors.
Candidate prognostic factor
Category
Participants
(n = 250)
Age (y), mean (SD) 50 (15)
Body mass index (kg/m
2
), mean (SD) 26.5 (5.3)
Smoking, n (%)
Never 166 (66)
Used to smoke but have quit 69 (28)
Current smoker 15 (6)
Exposure to heavy loads, n (%)
a
Rarely (rarely, very rarely or never) 100 (40)
Occasionally 84 (34)
Frequently (frequently or very frequently) 66 (26)
Exposure to awkward posture, n (%)
a
Rarely (rarely, very rarely or never) 110 (44)
Occasionally 74 (30)
Frequently (frequently or very frequently) 66 (26)
Physical activity, n (%)
Vigorous 124 (50)
Moderate 56 (22)
Low 70 (28)
Time sitting (hours), n (%)
a
0 to 5 111 (44)
.5 139 (56)
General health, n (%)
a
Excellent (excellent or very good) 128 (51)
Good 99 (40)
Poor (fair or poor) 23 (9)
Number of previous episodes of LBP, n (%)
a
1 to 2 70 (28)
3to10 93(37)
.10 87 (35)
Duration of last episode of LBP (weeks), n (%)
a
,2 weeks 146 (58)
2 to 6 weeks 51 (20)
.6 weeks 53 (21)
Perceived risk of recurrence of LBP (points), n (%)
a
0 to 5 125 (50)
.5 125 (50)
Depression, n (%)
a
Normal (normal or mild) 215 (86)
Moderate (moderate, severe or extremely severe) 35 (14)
Anxiety, n (%)
a
Normal (normal or mild) 206 (82)
Moderate (moderate, severe or extremely severe) 44 (18)
Stress, n (%)
a
Normal (normal or mild) 204 (82)
Moderate (moderate, severe or extremely severe) 46 (18)
Sleep quality, n (%)
a
Good (very good or fairly good) 175 (70)
Bad (fairly bad or very bad) 75 (30)
LBP = low back pain.
a
Coded as used in the analysis of multivariate model.
4da Silva et al: Risk of recurrence of low back pain
(completeness index of 96%). To investigate the cumulative
probability of a recurrence of LBP, we used a consensus definition
for recurrence of an episode of LBP
13
as the primary outcome;
however, because it is clear that recurrence estimates are affected
by the definition, two other definitions of recurrence were
investigated. Finally, a range of predictor variables for a
recurrence were chosen a priori and measured using previously
published
19,22–25
and validated questionnaires.
26,27
This study also had some limitations. First, although we enrolled
consecutive participants, we acknowledge that some potentially
eligible participants might not have been informed about the study
by their clinicians, and others could not be contacted. This issue
could have precluded potentially eligible individuals from partici-
pating in the study. Also, self-reported measures were used to collect
information about all prognostic factors. Because the baseline
assessment was conducted over the telephone, some objective
measures (eg, strength) were not possible to measure and may have
predicted recurrence. We did not collect details of interventions that
participants received during the previous episode of LBP, which may
influence the time to recurrence of future episodes. Finally, the es-
timates for recurrence were based upon a clinical sample and may
not apply to people who do not seek healthcare for an episode of LBP.
A systematic review, published in 2017, investigated the risk of and
prognostic factors for a recurrence of LBP in patients who had recovered
from a previous episode of LBP within the last year.
4
The systematic
review included eight studies; however, only one included study was
considered to have an adequately short inception period (,6 weeks).
6
Stanton et al reported a recurrence of an episode of LBP estimate of 24%
(95% CI 20 to 28) within 12 months, based on a 12-monthrecall period.
6
When recurrence was defined as recall of a recurrence at 12 months or
pain reported at the 3-month or 12-month follow-up (even if partici-
pants failed to report a recurrence at 12 months), the recurrence esti-
mate increased to 33% (95% CI 28 to 38).
6
That study
6
was a secondary
analysis of a cohort study
34
investigating prognosis in patients with
recent-onset LBP in Australian primary care, and was therefore not
primarily designed to investigaterisk of recurrence. Only two studies
5,6
included in the systematic review
4
presented data on prognostic
factors for a recurrence of LBP. A history of previous episodes of LBP
prior to the most recent episode was the only significant predictor of
recurrence of LBP in both included studies. Hancock et al also found that
disc degeneration and high intensity zone (from magnetic resonance
imaging scans) were predictive of a recurrence of an episode of LBP.
5
Stanton et al
6
found, in a secondary analysis (without the variable
previous episodes of LBP in the model), that perceived risk of persistent
pain was a significant predictor of recurrence of an episode of LBP at 12
months.
Machado et al reported an estimate of recurrence of an episode of
LBP of 33% based on 1-year recall and an estimate of recurrence of LBP
causing care-seeking of 18%.
28
That study
28
was a secondary analysis
of a case-crossover study
22
investigating triggers of an episode of
acute LBP. Using a 1-year recall period likely contributed to the lower
rates of recurrence compared with the current findings.
The findings of this study have important implications for
clinicians, patients and future research. The results demonstrate
that after recovery from an episode of LBP, about 70% of people
will experience another episode of LBP within 1 year; however,
many of these episodes appear to be relatively minor because
whentherecurrenceoccurred,only about 40% of participants
reported moderate activity limitation or that healthcare was
sought. These results are important for clinicians when providing
information to patients about the likely risk and nature of
recurrences of LBP. The results also demonstrate the need for
effective strategies to prevent a recurrence of LBP. A systematic
review investigating the effectiveness of interventions for
prevention of LBP found moderate-quality evidence that exercise
combined with education reduces the risk of an episode of LBP by
45% (95% CI 26 to 59).
35
This program may be particularly
Probability of an episode of LBP
Time to recurrence of LBP (days)
0 100 200 300 400
1
0.8
0.6
0.4
0
0.2
A
Probability of activity-limiting LBP
Time to recurrence of activity-limiting LBP (days)
0 100 200 300 400
1
0.8
0.6
0.4
0
0.2
B
Probability of LBP for which care is sought
Time to recurrence of LBP for which care was sought (days)
0 100 200 300 400
1
0.8
0
0.2
0.6
0.4
C
Figure 2. Kaplan-Meier survival curves based on uncorrected inception time of
recurrence of: (A) an episode of LBP; (B) activity-limiting LBP; and (C) LBP for which
care was sought (all n = 250).
LBP = low back pain.
Research 5
important in those patients who were identified to be at increased
risk. The prognostic factors we identified of exposure to awkward
postures and long periods of sitting provide potential targets for
the development of more effective prevention strategies,
which would need to be tested in future studies. However, our
study does not provide evidence of a causal relationship between
these variables and recurrences, or evidence on the potential
mechanisms that might produce a causal relationship. Further
research to understand the mechanisms producing recurrences of
LBP is important.
What was already known on this topic: Although most
people with low back pain recover quickly, recurrence of low
back pain is common. Most previous prognostic studies of
recurrence of low back pain have not enrolled participants at a
uniform time point after their previous episode of low back pain.
What this study adds: This study enrolled people who had
recovered from an episode of low back pain within the past
month, and followed them for 12 months. About 70% of people
in the study had a recurrence of low back pain during follow-up.
Prognostic factors for a recurrence include exposure to awkward
posture, longer time sitting, and more than two previous epi-
sodes of low back pain.
eAddenda: Table 4 can be found online at https://doi.org/10.1016/j.
jphys.2019.04.010.
Ethics approval: The Human Research Ethics Committee,
Macquarie University approved this study (#5201500494).
Table 5
Univariate analyses for a recurrence of an episode of LBP.
Candidate prognostic factor
Category
Recurrence (n/total n)
a
Hazard ratio 95% CI P
Age (y) N/A 1.00 0.99 to 1.01 0.38
Body mass index (kg/m
2
)N/A 1.01 0.98 to 1.04 0.65
Smoking history
Never smoked 118/166 Reference –0.57
Used to smoke but have quit 43/69 0.85 0.60 to 1.21 0.36
Current smoker 9/15 0.80 0.41 to 1.58 0.52
Exposure to heavy loads
Rarely (rarely, very rarely or never) 64/100 Reference –0.22
Occasionally 60/84 1.36 0.96 to 1.93 0.09
Frequently (frequently or very frequently) 46/66 1.26 0.86 to 1.84 0.23
Exposure to awkward posture
Rarely (rarely, very rarely or never) 70/110 Reference –0.07
Occasionally 52/74 1.23 0.86 to 1.77 0.25
Frequently (frequently or very frequently) 48/66 1.54 1.07 to 2.23 0.02
Physical activity
Vigorous 85/124 Reference –0.71
Moderate 35/56 0.90 0.61 to 1.33 0.59
Low 50/70 1.08 0.76 to 1.53 0.68
Time sitting (hours)
0to5 66/111 Reference ––
.5 104/139 1.39 1.02 to 1.90 0.04
General health
Excellent 83/128 Reference –0.19
Good 73/99 1.32 0.96 to 1.80 0.09
Poor 14/23 0.96 0.54 to 1.69 0.88
Number of previous episodes of LBP
1to2 36/70 Reference –0.002
3to10 67/93 1.67 1.11 to 2.50 0.01
.10 67/87 2.09 1.39 to 3.13 ,0.001
Duration of previous episode of LBP (weeks)
,2 98/146 Reference –0.22
2to6 39/51 1.33 0.92 to 1.93 0.13
.6 33/53 0.91 0.62 to 1.36 0.65
Perceived risk of recurrence of LBP (points)
0to5 77/125 Reference ––
.5 93/125 1.56 1.16 to 2.12 0.004
Depression
Normal (normal or mild) 144/215 Reference ––
Moderate (moderate, severe or extremely severe) 26/35 1.09 0.72 to 1.65 0.69
Anxiety
Normal (normal or mild) 140/206 Reference ––
Moderate (moderate, severe or extremely severe) 30/4 4 1.02 0.69 to 1.51 0.92
Stress
Normal (normal or mild) 136/204 Reference ––
Moderate (moderate, severe or extremely severe) 34/46 1.23 0.92 to 1.96 0.27
Sleep quality
Good (very good or fairly good) 115/175 Reference ––
Bad (fairly bad or very bad) 55/75 1.20 0.87 to 1.66 0.26
LBP = low back pain, N/A = not applicable.
a
Total = number of participants in category.
Table 6
Multivariate analysis for a recurrence of an episode of LBP.
Prognostic factor
Category
Hazard ratio 95% CI P
Exposure to awkward posture
Rarely (rarely, very rarely or never) Reference –0.01
Occasionally 1.20 0.83 to 1.73 0.34
Frequently (frequently or very frequently) 1.81 1.22 to 2.68 0.003
Time sitting (hours)
0 to 5 Reference ––
.5 1.50 1.08 to 2.09 0.02
Number of previous episodes of LBP
1 to 2 Reference –0.0 02
3 to 10 1.63 1.08 to 2.47 0.01
.10 1.94 1.28 to 2.94 ,0.0 01
LBP = low back pain.
6da Silva et al: Risk of recurrence of low back pain
All participants gave verbal informed consent before data collection
began.
Competing interests: Nil.
Sources of support: This work was supported by the Department
of Health Professions of Macquarie University. Tatiane da Silva has a
PhD scholarship from CAPES (Coordenação de Aperfeiçoamento de
Pessoal de Nível Superior), Ministry of Education of Brazil. Prof. Chris
Maher’s fellowship is funded by Australia’s National Health and
Medical Research Council. Tarcisio de Campos has a PhD scholarship
from Macquarie University (Macquarie University Research Excel-
lence Scholarship - MQRES).
Acknowledgements: We thank all the participants, physiothera-
pists and chiropractors who participated in the study.
Provenance: Not invited. Peer reviewed.
Correspondence: Dr Tatiane da Silva, Department of Health Pro-
fessions, Macquarie University, Sydney, Australia. Email:
Tatiane.mota-da-silva@students.mq.edu.au
References
1. da Cunha Menezes Costa L, Maher CG, Hancock MJ, McAuley JH, Herbert RD,
Costa LO. The prognosis of acute and persistent low-back pain: a meta-analysis.
CMAJ. 2012;184:E613–E624.
2. Refshauge KM, Maher CG. Low back pain investigations and prognosis: a review. Br
J Sports Med. 2006;40:494–498.
3. Hoy D, Brooks P, Blyth F, Buchbinder R. The epidemiology of low back pain. Best
Pract Res Clin Rheumatol. 2010;24:769–781.
4. da Silva T, Mills K, Brown BT, Herbert RD, Maher CG, Hancock MJ. Risk of
recurrence of low back pain: a systematic review. J Orthop Sports Phys Ther.
2017;47:305–313.
5. Hancock MJ, Maher CM, Petocz P, Lin CW, Steffens D, Luque-Suarez A, et al. Risk
factors for a recurrence of low back pain. Spine J. 2015;15:2360–2368.
6. Stanton TR, Henschke N, Maher CG, Refshauge KM, Latimer J, McAuley JH. After an
episode of acute low back pain, recurrence is unpredictable and not as common as
previously thought. Spine. 2008;33:2923–2928.
7. Biering-Sorensen F. A prospective study of low back pain in a general population. I.
Occurrence, recurrence and aetiology. Scand J Rehabil Med. 1983;15:71–79.
8. Carey TS, Garrett JM, Jackman A, Hadler N. Recurrence and care seeking after acute
back pain: results of a long-term follow-up study. North Carolina Back Pain Project.
Med Care. 1999;37:157–164 .
9. Cassidy JD, Cote P, Carroll LJ, Kristman V. Incidence and course of low back pain
episodes in the general population. Spine. 2005;30:2817–2823.
10. Stevenson JM, Weber CL, Smith JT, Dumas GA, Albert WJ. A longitudinal study of
the development of low back pain in an industrial population. Spine.
2001;26:1370–1377.
11. Herbert R, Jamtvedt G, Hagen K, Mead J. Practical Evidence-Based Physiotherapy.
2nd ed. Edinburgh, UK: Elsevier/Churchill Livingstone; 2011.
12. Laupacis A, Wells G, Richardson WS, Tugwell P. Users’guides to the medical
literature. V. How to use an article about prognosis. Evidence-Based Medicine
Working Group. JAMA. 1994;272:234–237.
13. Stanton TR, Latimer J, Maher CG, Hancock MJ. A modified Delphi approach to
standardize low back pain recurrence terminology. Eur Spine J. 2011;20:744–752.
14. Soukup MG, Glomsrod B, Lonn JH, Bo K, Larsen S. The effect of a Mensendieck
exercise program as secondary prophylaxis for recurrent low back pain.
A randomized, controlled trial with 12-month follow-up. Spine. 1999;24:
1585–1591.
15. Soukup MG, Lonn J, Glomsrod B, Bo K, Larsen S. Exercises and education as
secondary prevention for recurrent low back pain. Physiother Res Int. 2001;6:27–
39.
16. van Tulder M, Becker A, Bekkering T, Breen A, del Real MT, Hutchinson A, et al.
Chapter 3. European guidelines for the management of acute nonspecific low back
pain in primary care. Eur Spine J. 2006;15:S169–S191.
17. Hartvigsen J, Hancock MJ, Kongsted A, Louw Q, Ferreira ML, Genevay S, et al. What
low back pain is and why we need to pay attention. Lancet. 2018;391:2356–2367.
18. Maher C, Underwood M, Buchbinder R. Non-specific low back pain. Lancet.
2017;389:736–747.
19. Williams CM, Maher CG, Latimer J, McLachlan AJ, Hancock MJ, Day RO, et al. Effi-
cacy of paracetamol for acute low-back pain: a double-blind, randomised
controlled trial. Lancet. 2014;384:1586–1596.
20. Stevens ML, Lin CW, Hancock MJ, Latimer J, Buchbinder R, Grotle M, et al. TOPS:
Trial Of Prevention Strategies for low back pain in patients recently recovered from
low back pain-study rationale and protocol. BMJ Open. 2016;6:e011492.
21. McGinn TG, Guyatt GH, Wyer PC, Naylor CD, Stiell IG, Richardson WS. Users’guides
to the medical literature: XXII: how to use articles about clinical decision rules.
Evidence-Based Medicine Working Group. JAMA. 2000;284:79–84.
22. Steffens D, Ferreira ML, Latimer J, Ferreira PH, Koes BW, Blyth F, et al. What triggers
an episode of acute low back pain? A case-crossover study. Arthritis Care Res.
2015;67:403–410.
23. Armstrong T, Bauman AE, Davies J, Australian Institute of Health and Welfare.
Physical activity patterns of Australian adults: results of the 1999 national physical
activity survey. Canberra: Australian Institute of Health and Welfare; 2000.
24. da Silva T, Macaskill P, Mills K, Maher C, Williams C, Lin C, et al. Predicting recovery
in patients with acute low back pain: a clinical prediction model. Eur J Pain.
2017;21:716–726.
25. Hancock MJ, Maher CG, Latimer J, Herbert RD, McAuley JH. Can rate of recovery be
predicted in patients with acute low back pain? Development of a clinical pre-
diction rule. Eur J Pain. 2009;13:51–55.
26. Gloster AT, Rhoades HM, Novy D, Klotsche J, Senior A, Kunik M, et al. Psychometric
properties of the Depression Anxiety and Stress Scale-21 in older primary care
patients. J Affect Disord. 2008;110:248–259.
27. Ware JE, Jr., Sherbourne CD. The MOS 36-item short-form health survey (SF-36). I.
Conceptual framework and item selection. Med Care. 1992;30:473–483.
28. Machado GC, Maher CG, Ferreira PH, Latimer J, Koes BW, Steffens D, et al. Can
recurrence after an acute episode of low back pain be predicted? Phys Ther.
2017;97:889–895.
29. Peduzzi P, Concato J, Kemper E, Holford TR, Feinstein AR. A simulation study of the
number of events per variable in logistic regression analysis. J Clin Epidemiol.
1996;49:1373–1379 .
30. Peduzzi P, Concato J, Feinstein AR, Holford TR. Importance of events per inde-
pendent variable in proportional hazards regression analysis. II. Accuracy and
precision of regression estimates. J Clin Epidemiol. 1995;48:1503–1510.
31. Clark TG, Altman DG, De Stavola BL. Quantification of the completeness of follow-
up. Lancet. 2002;359:1309–1310.
32. Hancock MJ, Maher CG, Costa Lda C, Williams CM. A guide to survival analysis for
manual therapy clinicians and researchers. Man Ther. 2014;19:511–516.
33. Moons KG, Altman DG, Reitsma JB, Ioannidis JP, Macaskill P, Steyerberg EW, et al.
Transparent Reporting of a multivariable prediction model for Individual Prognosis
or Diagnosis (TRIPOD): explanation and elaboration. Ann Intern Med.
2015;162:W1–W73.
34. Henschke N, Maher CG, Refshauge KM, Herbert RD, Cumming RG, Bleasel J, et al.
Prognosis in patients with recent onset low back pain in Australian primary care:
inception cohort study. BMJ. 2008;337:a171.
35. Steffens D, Maher CG, Pereira LS, Stevens ML, Oliveira VC, Chapple M. Prevention of
low back pain: a systematic review and meta-analysis. JAMA Intern Med.
2016;176:199–208.
Research 7
