ORIGINAL ARTICLE

Interventions to address deficits of pharmacological pain

management in nursing home residents A cluster-randomized
trial
nner1, A. Budnick2, R. Kuhnert1, I. Wulff2, S. Kalinowski2, P. Martus3, D. Dra
ger2, R. Kreutz1
F. Ko
-Universitatsmedizin Berlin, Germany
1 Institute of Clinical Pharmacology and Toxicology, Charite
2 Institute of Medical Sociology and Rehabilitation Science, Charit
e- Universitatsmedizin Berlin, Germany
bingen, Germany
3 Institute for Clinical Epidemiology and Applied Biometry, University Hospital Tu

Correspondence
Reinhold Kreutz
E-mail: reinhold.kreutz@charite.de
Funding sources
This project was funded by the Federal Ministry of Education and Research of Germany
(grant number 01ET1001A).
Conflicts of interest
None declared.

Accepted for publication

19 December 2014
doi:10.1002/ejp.663

Abstract
Background: To evaluate the effect of interventions for general
practitioners and nursing home staff to improve pain severity and
appropriateness of pain medication in nursing home residents (NHR).
Methods: This cluster-randomized controlled trial was conducted in six
nursing homes in the intervention and control group, respectively. Pain
management was analysed before (T0) and after (T1, T2) an educational
intervention in 239 NHR, aged 65 years, without moderate or severe
cognitive impairment. Primary and secondary outcomes were average
pain severity and appropriateness of pain medication as determined with
the Numeric Rating Scale and Pain Medication Appropriateness Scale
(PMASD), respectively.
Results: At T0, 72.2% and 73.7% of NHR (mean age 83 years) reported
pain (average pain severity 2.4) in the intervention and control group,
respectively. The PMASD at T0 was 53.9 in the intervention group and 60.8
in the control group (p = 0.12), while 20.6% compared to 6.9%
(p = 0.009) received no pain medication in the two groups. At T2, nonsignificant improvements in the average pain severity (1.59) and PMASD
(61.07) were observed in the intervention group. Moreover, the mean
individual PMASD increased by 8.09 (p = 0.03) and the proportion of NHR
without pain medication decreased by 50% (p = 0.03) in the intervention
group. No appreciable changes were found in the control group at T2.
Conclusions: NHR exhibited a high prevalence of pain with overall low
severity, while a high proportion of individuals received inappropriate
pain medications. Both findings were not significantly improved by the
intervention, although some aspects of drug treatment were meaningful
improved.

1. Introduction
Pain in nursing home residents (NHR) remains an
important research topic since deficits in pain management were revealed in several studies (Won et al.,
2004; Drager et al., 2013; K
olzsch et al., 2012; Lapane
2015 European Pain Federation - EFIC

et al., 2013). Up to 84% of NHR are affected by pain

worldwide (Takai et al., 2010) and about 50% in Germany (Dr
ager et al., 2013; Osterbrink et al., 2012;
Lukas et al., 2013). Previous studies indicated that
about 25% of NHR receive no pain medication
although they have pain (Won et al., 2004; K
olzsch
et al., 2012; Lapane et al., 2013). Persistent or
Eur J Pain 19 (2015) 1331--1341

1331

nner et al.
F. Ko

Appropriate pain management in nursing home residents

Whats already known about this topic?

Deficits in pain management in nursing home
residents have been reported in several countries including Germany.
What does this study add?

This is the first intervention study in Germany

addressing general practitioners and nursing
home staff to improve pain management.
A pain medication appropriateness scale was
applied longitudinally.
The findings indicate that the use of multifaceted educational interventions is a first step
to improve pain management and pain medication use in nursing home residents.

untreated pain is related to several consequences such

as functional impairment, depression, falls or sleep disorders, which can eventually lead to a reduction in
quality of life (The American Geriatrics Society, 2002).
The therapeutic approach of pain treatment in older
adults differs from treatment of younger patients, not
only because of altered drug effects but also because
the superior goal in the elderly is to maintain and
improve quality of life. This is a challenging task for
general practitioners (GPs), and requires individual
medication decisions based on weighing the risks and
benefits (The American Geriatrics Society, 2002). Several clinical guidelines to enhance pain management
in older adults have been developed and disseminated
(The American Geriatrics Society, 2002, 2009; Pergolizzi et al., 2008; Abdulla et al., 2013), and several
studies were conducted to improve pain management
in nursing homes (Herman et al., 2009), but
approaches in Germany are lacking.
In the context of this study, an interdisciplinary
guidance for pain management in nursing home residents (Wulff et al., 2012) was developed which
specifically addresses all professions who assess and
treat pain in NHR: GPs, pharmacists, nurses and
administrators of nursing homes. In the presented
trial, this guidance was used to develop multifaceted
educational programs for GPs and nursing home
staff. The aim of the current study was to compare
the effect of these programs regarding the improvement of the average pain severity and the appropriateness of pain medication use between NHR who
were treated by intensively trained GPs and nursing
home staff and NHR who were treated by GPs and
nursing home staff who received basic training.

1332 Eur J Pain 19 (2015) 1331--1341

2. Methods
2.1. Setting and study design
This study was conducted in 12 nursing homes in
Berlin, Germany, which were provided by one forprofit chain. Written consent was received by the
senior manager. The design of a cluster-randomized
controlled trial was chosen to avoid a contamination
of the interventional effect as the interventions were
conducted on the level of the nursing home (=cluster) (Wears, 2002). Hence, individual randomization
of GPs or NHR appeared inappropriate. The allocation sequence was generated by an experienced statistician, and the complex enrolment of participants
was conducted by two experienced researchers of
the study team. Six nursing homes were randomly
assigned to either the intervention group (IG) or
control group (CG). The remaining six nursing
homes had to be paired (4:2) before random assignment because the involved GPs of one practice treated NHR in four nursing homes and GPs of another
practice cared for NHR in two nursing homes. The
majority of the participating GPs was working in
independent private practices and thus not employed
by the owner of the nursing home.
Based on previous data (Dr
ager et al., 2013;
K
olzsch et al., 2012), a sample size of 96 NHR in
each arm was calculated to detect a difference of 2.0
points (SD 3.0) (Wood et al., 2010) on the Numeric
Rating Scale (NRS) between the IG and the CG. This
calculation considered a power of 80%, a two-sided
alpha of 0.05, and an estimated intraclass correlation
coefficient (ICC) of 0.05. The calculation revealed
that an ICC of 0.12 would still maintain the power
of 80%. To enrol this number of subjects, it was estimated to screen about 1500 NHR taking into account
a follow-up mortality rate of 10%, a prevalence of
50% for cognitive status (Mini Mental State Examination (MMSE) score (Folstein et al., 1975) 18)
and a pain prevalence of 55%.
The study is registered at the German section of
the International Clinical Trials Registry Platform of
the
World
Health
Organization
(DRKS-ID:
DRKS00004239); it complies with the Declaration of
Helsinki, and was approved by the local ethics committee (EA2/150/11).

2.2. Data collection

Data were collected between February 2012 and
December 2012. Personal interviews with NHR and
the analysis of the nursing home medical records

2015 European Pain Federation - EFIC

nner et al.
F. Ko

were conducted by trained interviewers at baseline

(T0), after three (T1) and six (T2) months. A computer-assisted full medication inventory was completed using the Instrument for Database-assisted
Online recording for Medication IDOM (M
uhlberger et al., 2003). All drugs were categorized according to the Anatomical Therapeutic Chemical (ATC)
Classification. Pain medication included drugs classified as ATC N02 (Analgesics) and M01A (Antiinflammatory and Antirheumatic Products, Non-Steroids) (WHO, 2012). The information about prescribed drugs including dosing and diagnoses were
obtained from the nursing homes medical records.

2.3. Study population

NHR were eligible to participate who were at least
65 years old, spent more than 3 months in the facility and had no or mild cognitive impairment (MMSE
Score (Folstein et al., 1975) 18). Written informed
consent was obtained from the residents or their
legal representatives before data collection.

2.4. Interventions
Two multifaceted education programs were developed by the interdisciplinary study team for the residents GPs and nursing home staff. Both programs
were based on an interdisciplinary guidance (Wulff

Appropriate pain management in nursing home residents

et al., 2012) and focused on knowledge (basics of

pain, pain assessment, non-drug therapy and drug
therapy) and reflecting personal attitudes towards
pain management in NHR and the collaboration
between GPs and nurses (Figs. 12). This approach
was chosen because previous studies reported that a
lack of knowledge and poor attitudes represent
major barriers for providing adequate pain management in nursing homes (Tarzian and Hoffmann,
2004). It is further supported by findings which suggest that self-awareness is the key in changing
behaviour (Hicks, 2000) and knowledge itself does
not easily lead to a change in action (Nazareth et al.,
2002).
All involved GPs were informed about the characteristics of the study prior to baseline data collection.
Subsequently, an online course, requiring an estimated working time of 56 h, was offered for
6 weeks after T0 to the GPs of the IG, while the GPs
of the CG could participate after the data collection
was completed. Furthermore, GPs received reminders and a printed short version of the interdisciplinary guidance. Participants received Continuing
Medical Education (CME) points after completion of
the course. Of 92 GPs, 22 (23%) agreed to participate in the trial. Fourteen GPs practised in the IG
and nine (64%) participated in the online course,
while two of eight (25%) GPs of the CG participated.

Figure 1 Interactive online course for general practitioners with exercises and assessment of training success.

2015 European Pain Federation - EFIC

Eur J Pain 19 (2015) 1331--1341

1333

nner et al.
F. Ko

Appropriate pain management in nursing home residents

Figure 2 Interactive, multi-perspective workshop for nursing home staff with handouts *was also content of 45-min basic training.

For nursing home staff of the IG facilities, 1-day

seminars (360 min) were conducted on-site by the
nurse and pharmacist of the research team. Nursing
home administrators were also encouraged to attend.
The seminar included presentations, exercises and
the opportunity for mutual exchange. Participants
received printed material (advantages and disadvantages of analgesic drugs for the elderly, pain assessment strategies and non-drug approaches). Eleven
seminars were conducted and 100 nursing home
staff members (18 were members of management
level) participated (79%). Nursing home staff of each
CG facility was offered a 45-min presentation about
general pain management and 69 (60%) nurses
participated.

2.5. Measures
2.5.1. Pain
Pain assessment was initiated by two dichotomous
questions are you in pain? and are you occasionally in pain? and by applying the Brief Pain Inventory (BPI) (Radbruch et al., 1999). Average pain
severity was determined with the BPI as the score
on NRS. Persistent pain was defined as reporting
pain in at least two measurements.
The primary outcome was the reported average
pain severity on the NRS within the last 24 h.

olzsch et al., 2013) and used in a non(PMASD) (K

interventional cross-sectional study (K
olzsch et al.,
2012).
The original PMAS tool and its German version
PMASD consist of 10 criteria and evaluate several
aspects of pain treatment in NHR (Table 3). This
includes the assessment of the type of prescribed
pain medication, dosages and dosing intervals for
certain pain syndromes, laxative co-medication during opioid therapy and the prescription of high-risk
drugs for the elderly. In addition to this objective
evaluation, the subjective perception of pain severity
and therapeutic success of the individual is also part
of the PMAS evaluation. An extensive training for
the evaluation of the PMASD tool was conducted
prior to data analysis that included an introduction
course in the use of the PMASD tool and the determination of the inter-rater reliability in a random
sample of 71 individuals.
The PMASD was applied once at each measuring
point to NHR who reported pain according to the
BPI and/or received pain medication. PMASD values
<67 indicate an inappropriate pain medication strategy according to the original authors (Hutt et al.,
2006). Although the sample size calculation was initially based on the NRS, this study had a power of
80% to detect between- and within-group differences of 10% and 7.5% in PMASD, respectively.

2.5.2. Appropriateness of pain medication

2.6. Data analyses

The Pain Medication Appropriateness Scale (PMAS)

(Hutt et al., 2006) was used to evaluate the secondary outcome, i.e. the mean PMAS values. The PMAS
was previously adapted to the German setting

The PMASD score was calculated as described by Hutt

et al. (2006) and differences in the individual PMASD
scores were calculated as PMASDT2 PMASDT0. For
the analyses of differences between the IG and the

1334 Eur J Pain 19 (2015) 1331--1341

2015 European Pain Federation - EFIC

nner et al.
F. Ko

CG, a linear mixed-effect model was used with the

differentiation in IG and CG as fixed effect, and the
nursing home as random effect. Differences in categorical variables between the two groups were calculated with Pearsons v2 with Rao & Scott adjustment.
These calculations were performed in R (package
nlme, version 3.1109 and package survey). Differences between measuring points were calculated with
McNemars Test using IBM SPSS Statistics for Windows, Version 19.0. (IBM Corp., Armonk, NY, USA).
Statistical significance was determined with an alpha
of 0.05.
Intraclass correlation coefficients (ICC) were calculated as described by Smeeth and Ng (2002).

3. Results
A total of 747 NHR fulfilled the inclusion criteria
(Fig. 3), and 239 (70% female, mean age 82.9,
8.1 years) NHR were studied at T0, while 206 and
177 were studied at T1 and T2 follow-up. Seven
interviews were discontinued and 16 deaths (6.7%)
occurred between T0 and T2. At T0, the study
groups did not significantly differ according to age,
sex, MMSE score or pain-associated diagnoses,
except osteoporosis (p = 0.007, Table 1). Furthermore, there were no differences regarding the proportion of NHR reporting pain or the mean average
pain severity on NRS (Table 2). However, at T0,
20.6% of NHR received no pain medication in IG
compared to 6.9% in CG (p = 0.009, Table 2). The
mean PMASD values were not significantly different,
although numerically lower (53.91 vs. 60.85) in IG
compared to CG (Table 2). All medication appropriateness criteria are presented in Table 3.
At T2, the proportion of NHR with persistent pain
was similar in the IG and CG (80.6% vs. 86.4%,
Table 2), while the mean average pain severity
tended to be lower in the IG compared to the CG
(1.59 vs. 2.45, p = 0.07, Table 2). The number of
NHR without pain medication use was significantly
reduced from 20.6% to 10.3% in the IG (p = 0.03),
while an increase, albeit small, from 6.9% to 8.7%
was observed in the CG (p = 0.62).
The inter-rater reliability of the PMASD was high
(r = 0.96, p < 0.001) and comparable to previous
studies (Hutt et al., 2006; K
olzsch et al., 2012). The
analysis of the mean PMASD values at T2 revealed a
non-statistically significant increase of 7.2 in the IG,
while the mean PMASD in the CG remained
unchanged ( 0.4, Table 2). Differences in the
PMASD values between T0 and T2 of each individual
NHR were calculated as a parameter reflecting longi 2015 European Pain Federation - EFIC

Appropriate pain management in nursing home residents

tudinal changes. This calculation demonstrated a

statistically significant mean PMASD increase of 8.09
within the IG (p = 0.03) and a non-significant
increase of 3.09 within the CG (p = 0.37), respectively. No statistically significant changes of any single PMASD criteria between T0 and T2 were
observed although some numerical differences in
single items were recorded (Table 3). Overall, only
small changes from T0 to T2 of mean PMASD values
were observed within each individual nursing home
in both groups (Fig. 4). However, one facility of the
IG showed a profound increase of 23.0 between T0
and T2, which was significantly greater as compared
to the mean increase in PMASD in the CG
(p = 0.02). Furthermore, in this nursing home, the
proportion of NHR with PMASD <67 significantly
decreased from 71% at T0 to 33% at T2 (p = 0.05).
In overall 25 subjects of the IG, a new analgesic
drug was started after T0. Of those NHR, 17 individuals received an additional on demand [pro re nata
(PRN)] medication, 7 a scheduled medication and 1
NHR received an additional PRN plus scheduled medication. Dipyrone (Metamizole) was the most frequent
additionally prescribed analgesic drug (n = 10) and
was in most cases (80%) prescribed as PRN. New opioids were also additionally prescribed at T1 and T2,
respectively, of which in four cases (57.1%), the individual did not receive an opioid at T0. In two cases,
the opioid was switched to another opioid. This
included a switch from tilidine to an oxycodone plus
naloxone combination and between oxycodone to
fentanyl. All NHR who additionally received an opioid
did not receive a laxative drug. Non-opioids such as
ibuprofen (n = 1, scheduled, n = 2 PRN) and etoricoxib (n = 1, scheduled), and acetaminophen (n = 1,
PRN) were also additionally prescribed.

4. Discussion
This is the first study in Germany involving GPs and
nursing home staff to address the deficits in pain
treatment in NHR. Furthermore, the PMASD was
applied to NHR longitudinally for the first time. The
prevalence of pain observed in this study was higher
than in other studies conducted in Germany (Drager
et al., 2013; Osterbrink et al., 2012). However, the
mean average pain severity at baseline was 2.4 and
thus lower than expected in the power calculation.
Consequently, the aimed reduction in two points in
pain severity on the NRS would result in a complete
absence of pain. In this regard, the observed reduction to a severity level of 1.59 in the IG at T2 may
still represent a reasonable result. Nevertheless,
Eur J Pain 19 (2015) 1331--1341

1335

Appropriate pain management in nursing home residents

nner et al.
F. Ko

Figure 3 Diagram showing flow of clusters and nursing home residents through the cluster-randomized trial according to Campbell et al. (Campbell et al., 2004). NH, nursing home; NHR, nursing home resident; GP, general practitioner; MMSE, Mini Mental State Examination. apart I: seminar
on pain management in nursing homes for nursing staff; part II: online course for GPs; part III: short presentation for nursing staff; part IV: short
presentation at recruiting meeting for GPs.

almost two-thirds of NHR were affected by inappropriate pain medication use indicating the need for
improved pain treatment in agreement with other
recent studies (Hutt et al., 2006; Dr
ager et al., 2013;
K
olzsch et al., 2012). At T0, PMASD values in the IG
were numerically lower compared to the CG. This
observation was related to the higher proportion of
NHR that did not receive any pain medication,
although they reported pain.
1336 Eur J Pain 19 (2015) 1331--1341

Interestingly, the overall increase in the mean

PMASD in the IG was largely driven by a profound
increase in one nursing home facility, while the
effect in the other nursing homes of the IG was
rather low (Fig. 4). This is due to the fact that a
large proportion of NHR in this particular nursing
home received additional pain medication. Furthermore, this facility was staffed with a nurse who
acquired the additional qualification as pain nurse
2015 European Pain Federation - EFIC

nner et al.
F. Ko

Appropriate pain management in nursing home residents

Table 1 Baseline characteristics of study population.

Mean age [95% CI]

Female sex n (%)
Mean MMSE [95% CI]
Pain associated diagnoses n (%)
Chronic pain
Osteoarthritis
Osteoporosis
Polyneuropathy

Intervention group (n = 121)

Control group (n = 118)

p-value

83.58 [81.2385.93]
86 (71.1)
23.47 [22.7924.15]

82.5 [80.1184.88]
81 (68.6)
23.89 [23.1624.61]

0.54a
0.70b
0.43a

5
10
11
7

(4.1)
(8.3)
(9.1)
(5.8)

8
12
22
8

ICC
0.077
0.004

0.25b
0.51b
0.007b
0.76b

(6.8)
(10.2)
(18.6)
(6.8)

MMSE, Mini Mental State Examination; CI, confidence interval; ICC, Intraclass correlation coefficient, a negative ICC results from a greater variance
within the cluster compared to the variance between the clusters.
a
Linear mixed-effect model.
b
Pearsons v2 with Rao & Scott adjustment.
Table 2 Prevalence of pain, mean average pain, analgesic drug use and appropriateness of pain medication.
T0

Total sample
Affected by paina,
n (%)
Reporting pain,
n (%)
Persistent painb,
n (%)
Mean average
painc [95% CI]
PMASD sampled

T2

Intervention
group
n = 121

Control
group
n = 118

p-value

ICC

Intervention
group
n = 89

Control
group
n = 88

p-value

100 (82.6)

95 (80.5)

0.73e

73 (82.0)

72 (81.8)

0.98e

88 (72.7)

87 (73.7)

0.88e

58 (65.9)

62 (70,5)

0.43e

50 (80.6)

57 (86.4)

0.37e

2.43 [1.892.97]

n = 92

Pain medication n (%)

Scheduled
16
PRN
32
Both
25
None
19
Mean PMASD
53.91
Score [95% CI]
PMASD score
58
<67, n (%)
Mean differences

of PMASD
score T0 to
T2 [95% CI]

(17.4)
(34.8)
(27.2)
(20.6)
[48.3759.45]
(63.0)

2.46 [1.93.03]

0.94f

0.02

1.59 [0.992.19]

2.45 [1.873.03]

n = 87

n = 68

n = 69

7
38
36
6
60.85

11
25
25
7
61.07

6
28
29
6
60.47

(8.0)
(43.7)
(41.4)
(6.9)
[55.1566.54]

53 (60.9)

0.009e,
0.11f

0.01

0.73e

(16.2)
(36.8)
(36.8)
(10.3)
[54.2967.85]

39 (57.3)

8.09 [0.9615.22]

(8.7)
(40.6)
(42.0)
(8.7)
[53.7167.23]

39 (56.5)
3.09 [ 3.76 to 9.95]

0.07f

0.01

0.80e,g
0.90f

0.01

0.92e
0.34f

CI, confidence interval; ICC, intraclass correlation coefficient; PRN, pro re nata; PMASD, Pain Medication Appropriateness Scale; range:
(highly inappropriate) to 100.0 (very appropriate).
a
Affected by pain was defined as reporting pain and/or receiving pain medication.
b
Persistent pain was defined as reporting pain in at least two measuring points.
c
Mean average self-reported pain severity in the last 24 h on Numeric Rating Scale (NRS, range 010).
d
PMASD was applied when NHR reported pain in BPI and/or received pain medication.
e
Pearsons v2 with Rao & Scott adjustment.
f
Linear mixed-effect model.
g
Compared any medication use to no medication use.

during the time this trial was conducted and was

exclusively responsible for general pain management. This result is supported by the fact that the
2015 European Pain Federation - EFIC

ICC

0.01

16.7

GPs who practised in this nursing home also practised in another nursing home facility in which
the appropriateness of pain medication did not
Eur J Pain 19 (2015) 1331--1341

1337

nner et al.
F. Ko

Appropriate pain management in nursing home residents

Table 3 PMASD criteria at T0 and T2.

T0

T2
Control
group (%)

Between
group
p-valueb

Intervention
group (%)

Control
group (%)

Between
group
p-valueb

Intra
group
p-value IGc

PMAS criteriona

Intervention
group (%)

1.
2.
3.
4.
5.
6.
7.

179
21
81
25
98
175
116

58.7
83.3
61.6
30.0
33.3
47.2
62.3

75.9
80.0
69.1
60.0
41.5
53.5
60.0

0.05
0.88
0.56
0.15
0.30
0.38
0.78

137
21
50
20
70
127
83

70.6
33.3
70.0
20.0
40.0
66.1
60.0

76.8
77.8
56.7
50.0
37.5
60.0
60.5

0.35
0.13
0.43
0.07
0.84
0.50
0.96

0.06
0.22
0.45
1.0
0.34
0.08
1.0

46

55.6

64.3

0.55

29

50.0

60.0

0.70

1.0

179

47.8

46.0

0.85

137

47.8

43.5

0.67

0.77

PRN for mild episodic pain

Neuropathic pain
Persistent pain
Recurrent pain
Breakthrough pain
Match pain severity
Degree of Pain
Relief from Medication
8./9. Appropriate Constipation
Prevention
10. Appropriate Exclusion of
Geriatric High-Risk Drugs

PRN, Pro re nata (as needed); Results imply appropriate prescribing (Hutt et al., 2006), i.e. Item 16: scores = 3; item 7: scores = 2 and 3; item 8/
9: scores 1, 2 and 3; item 10: score = 0.
a
Criterion was evaluated if applicable.
b
Pearsons v2 with Rao & Scott adjustment.
c
McNemars test.

Figure 4 PMASD values of all nursing home facilities at T0, T1 and T2. IG, Intervention group; CG, Control group. aNHR were treated by the same
GPs.

significantly improve (Fig. 4). Therefore, it may on

the one hand be assumed that the observed profound improvement may result from a synergistic
interaction between the intervention implemented
in this study and the personal skills of the pain nurse
1338 Eur J Pain 19 (2015) 1331--1341

resulting in a potentiation effect. Although this

observation may reflect a chance finding, it nevertheless has potential implications for further investigations to explore whether the effect of a pain nurse
is superior compared to the general effort of all
2015 European Pain Federation - EFIC

nner et al.
F. Ko

nurses. On the other hand, it may be speculated that

not only the education measures applied to GPs and
nursing home staff alone are important for a successful outcome but also the cooperation and communication between both. Although the latter aspects
were largely addressed in both intervention strands,
a joint seminar with GPs and nursing home staff participating together, to encourage a mutual inter-professional exchange, may have had a greater impact.
To test the effect of these strategies in further investigations appears thus a reasonable approach.
The mean average pain was mild in the population. Therefore, it has to be noted that, depending
on the residents therapy goals, it may not be treated
with medication. However, appropriate pain management in NHR is a challenging task. A large proportion of NHR was affected by polypharmacy (i.e.
total number of medication ranged from 1 to 22,
mean number of drugs 9.8) exposing them to an
increased risk for drugdrug interactions and additionally prescribed analgesics may result in an even
more complicated therapy scheme. The PMASD evaluates several aspects of pain treatment and revealed
deficits that mainly resulted from a lack of medication and/or use of inappropriate drugs. The additional prescribed analgesics in the IG after T0 were
not associated with drugdrug interactions or contraindications, i.e. renal insufficiency, but resulted in
an increase in the proportion of NHR with appropriate treatment regarding PMASD criteria 1 (PRN), 3
(persistent pain), 5 (breakthrough pain) and 6 (pain
severity). However, since further alterations in drug
prescribing patterns of GPs were scarce, there were
no changes regarding the remaining prescription of
high-risk drugs (criterion 10) or the lack of mandatory laxative co-medication during opioid therapy in
NHR (criterion 8 & 9), which has also been reported
in a previous study (K
olzsch et al., 2012). This also
explains the overall small improvements regarding
the proportion of NHR with a PMASD value <67 as
the recommended threshold of appropriate pain
medication (Hutt et al., 2006). These observations
imply that although new treatments were started
and the prescribed substances were generally chosen
carefully, there was still a potential for optimization
of drug treatment during the course of our study.
Dipyrone was the most frequently prescribed drug
throughout this population which is consistent with
another study conducted in German NHR (K
olzsch
et al., 2012). A permanent, scheduled prescription of
this drug is a matter of controversy (Ibanez et al.,
2005). Although dipyrone is not classified as a
potential inappropriate medication in the German
2015 European Pain Federation - EFIC

Appropriate pain management in nursing home residents

list of inappropriate medications in the elderly (Holt

et al., 2010), the risk for older and potentially frail
adults to develop harmful side effects, such as agranulocytosis, needs to be frequently monitored.
During this study, several challenges occurred,
which influenced the participation of nursing home
staff and GPs in the trial and the interventions,
respectively. Several organizational problems in the
nursing home facilities, such as staff shortages, hindered the participation in the seminar of several
members of the nursing home staff. Furthermore,
the proportion of temporarily employed nurses, who
did not participate in the seminars, was rather high.
Although the participation rate of the GPs in this
study was low (23%), it is important to notice that
these GPs cared for overall 71% of the NHR population. Reasons for the observed low participation rate
of GPs may have been their general high workload,
clinical inertia, a lack of awareness/knowledge of the
potential deficits in pain management or even the
absence of incentive payments (Miles, 2007; Hummers-Pradier et al., 2008). Furthermore, GPs may
have preferred to participate in an intervention with
training sessions including the possibility of mutual
exchange or networking with other GPs, pharmacists
or nurses, respectively. In contrast, an alternative
web-based type of intervention was chosen, because
it seemed more convenient for the GPs to participate
and in which CME points could be obtained. The
fact that the majority of GPs participating in this
study did not use this intervention represents a
major limitation and potential contributor to the
small improvements in pain management observed
in this study.
Limitations have to be noticed. The observed average pain severity was relatively low and the
observed changes potentially affected by floor effects.
Although the PMASD has been described as a reliable tool (Hutt et al., 2006; K
olzsch et al., 2012), the
recommended value of 67 as the threshold of inappropriate pain management has not been validated
and needs further investigation. Since the trial was
conducted in one for-profit chain, the results cannot
be generalized for all NHR in Germany. Furthermore, it has to be considered that NHR who were
repeatedly interviewed about their pain situation
might have been more conscious to ask nurses and
GPs for analgesics which may consequently lead to a
higher appropriateness in pain medication. This may
also explain the improvements, which were noticeable in individuals of the CG. Generally, the nursing
home population includes a high proportion of NHR
with cognitive impairment, but only NHR with no or
Eur J Pain 19 (2015) 1331--1341

1339

nner et al.
F. Ko

Appropriate pain management in nursing home residents

low cognitive impairment were included in this trial.

This strategy was chosen to test if the types of interventions would be overall successful, as the primary
outcome (reducing pain severity as estimated by a
quantitative scale) would not have been applicable
for cognitive impaired NHR and would have been
more difficult to measure in this population. Furthermore, the use of specific and appropriate pain
assessment tools in NHR with cognitive impairment
(Basler et al., 2006) would have required the inclusion of nursing home staff in the data collection procedure. This appeared too difficult, particularly
because of the limited personal resources currently
available in German nursing home.
In summary, in the current study, we observed a
high prevalence of pain at a low severity level and a
high proportion of individuals with inappropriate
pain medication. Some deficits in drug treatment,
e.g. the undertreatment of NHR, were noticeably
reduced; however, main outcomes were not statistically significant improved by the intervention. Further studies may aim to test the effectiveness of
interventions in NHR by using more intensive intervention tools such as real time prompts and alerts,
conducting an intervention for nursing home staff
and GPs together to strengthen their collaboration or
even the assignment of a pain nurse that may
improve the outcome in combination with educational efforts implemented in the current study.
Acknowledgements
We are grateful to the Helmholtz Zentrum M
unchen,
GmbH for licensing the IDOM database and the AOK
Research Institute (WIdO) for licensing the German Drug
Index. The authors thank Laura- Maria Jordan, Bianca
Hannemann, Andrea Woydack and Angela Kissel-Kr
oll for
their commitment and assistance in this project.

Author contributions
F.K. led GP part of the intervention and co-led nurse part
of intervention, conducted data ascertainment and analyses
of medication and documentation, led interpretation of the
data, primary manuscript author. A.B. led study design
and subject acquisition, coordinator of the trial, participated in analyses of the data and preparation of the manuscript. Ro.K. calculated and analysed data and participated
in preparation of the manuscript. I.W. led nurse part of
intervention and co-led GP part of the intervention,
participated in preparation of the manuscript. S.K. participated in preparation of the manuscript. P.M. conducted
the sample size calculation and participated in preparation
of the manuscript. D.D. and R.K. are principal investigators

1340 Eur J Pain 19 (2015) 1331--1341

responsible for the final study design and data analysis.

They also participated in preparation of the manuscript.

References
Abdulla, A., Adams, N., Bone, M., Elliott, A.M., Gaffin, J., Jones, D.,
Knaggs, R., Martin, D., Sampson, L., Schofield, P. (2013). Guidance
on the management of pain in older people. Age Ageing 42(Suppl 1),
i1i57.
Basler, H.D., Huger, D., Kunz, R., Luckmann, J., Lukas, A., Nikolaus,
T., Schuler, M.S. (2006). Assessment of pain in advanced dementia.
Construct validity of the German PAINAD. Schmerz 20, 519526.
Campbell, M.K., Elbourne, D.R., Altman, D.G. (2004). CONSORT
statement: Extension to cluster randomised trials. BMJ 328, 702708.
Drager, D., K
olzsch, M., Wulff, I., Kalinowski, S., Ellert, S., Kopke, K.,
Fischer, T., Kreutz, R. (2013). Autonomy despite of pain? Resources
and barriers in the life worlds of nursing home residents. In Kuhlmey,
A.
Tesch-R
omer,
C.,
ed.
(G
ottingen:
Autonomie
trotz
Multimorbiditat.
Ressourcen
f
ur
Selbststandigkeit
und
Selbstbestimmung im Alter. Hogrefe) 165202.
Folstein, M.F., Folstein, S.E., McHugh, P.R. (1975). Mini-mental
state. A practical method for grading the cognitive state of patients
for the clinician. J Psychiatr Res 12, 189198.
Herman, A.D., Johnson, T.M., Ritchie, C.S., Parmelee, P.A. (2009). Pain
management interventions in the nursing home: A structured review
of the literature. J Am Geriatr Soc 57, 12581267.
Hicks, T.J. (2000). Ethical implications of pain management in a
nursing home: A discussion. Nurs Ethics 7, 392398.
Holt, S., Schmiedl, S., Thurmann, P.A. (2010). Potentially inappropriate
medications in the elderly: The PRISCUS list. Dtsch Arztebl Int 107,
543551.
Hummers-Pradier, E., Scheidt-Nave, C., Martin, H., Heinemann, S.,
Kochen, M.M., Himmel, W. (2008). Simply no time? Barriers to GPs
participation in primary health care research. Fam Pract 25, 105112.
Hutt, E., Pepper, G.A., Vojir, C., Fink, R., Jones, K.R. (2006). Assessing
the appropriateness of pain medication prescribing practices in
nursing homes. J Am Geriatr Soc 54, 231239.
Ibanez, L., Vidal, X., Ballarin, E., Laporte, J.R. (2005). Agranulocytosis
associated with dipyrone (metamizol). Eur J Clin Pharmacol 60, 821
829.
K
olzsch, M., K
onner, F., Kalinowski, S., Wulff, I., Drager, D., Kreutz,
R. (2013). Quality and appropriateness of pain medication:
Instrument for estimation in nursing home residents. Schmerz 27,
497505.
olzsch, M., Wulff, I., Ellert, S., Fischer, T., Kopke, K., Kalinowski, S.,
K
Drager, D., Kreutz, R. (2012). Deficits in pain treatment in nursing
homes in Germany: A cross-sectional study. Eur J Pain 16, 439446.
Lapane, K.L., Quilliam, B.J., Chow, W., Kim, M.S. (2013).
Pharmacologic management of non-cancer pain among nursing
home residents. J Pain Symptom Manage 45, 3342.
Lukas, A., Mayer, B., Fialova, D., Topinkova, E., Gindin, J., Onder, G.,
Bernabei, R., Nikolaus, T., Denkinger, M.D. (2013). Pain
characteristics and pain control in European nursing homes: Crosssectional and longitudinal results from the Services and Health for
Elderly in Long TERm care (SHELTER) study. J Am Med Dir Assoc 14,
421428.
Miles, R.W. (2007). Fallacious reasoning and complexity as root causes
of clinical inertia. J Am Med Dir Assoc 8, 349354.
M
uhlberger, N., Behrend, C., Stark, R., Holle, R. (2003).
Datenbankgest
utzte Online-Erfassung von Arzneimitteln im Rahmen
gesundheitswissenschaftlicher Studien Erfahrungen mit der IDOMSoftware. Inform Biom Epidemiol Med Biol 34, 601611.
Nazareth, I., Freemantle, N., Duggan, C., Mason, J., Haines, A. (2002).
Evaluation of a complex intervention for changing professional
behaviour: The Evidence Based Out Reach (EBOR) Trial. J Health
Serv Res Policy 7, 230238.
Osterbrink, J., Hufnagel, M., Kutschar, P., Mitterlehner, B., Kruger, C.,
Bauer, Z., Aschauer, W., Weichbold, M., Sirsch, E., Drebenstedt, C.,

2015 European Pain Federation - EFIC

nner et al.
F. Ko

Perrar, K.M., Ewers, A. (2012). The pain situation for residents of

nursing homes. Results of a study in Munster. Schmerz 26, 2735.
Pergolizzi, J., Boger, R.H., Budd, K., Dahan, A., Erdine, S., Hans, G.,
Kress, H.G., Langford, R., Likar, R., Raffa, R.B., Sacerdote, P. (2008).
Opioids and the management of chronic severe pain in the elderly:
Consensus statement of an International Expert Panel with focus on
the six clinically most often used World Health Organization Step III
opioids (buprenorphine, fentanyl, hydromorphone, methadone,
morphine, oxycodone). Pain Pract 8, 287313.
Radbruch, L., Loick, G., Kiencke, P., Lindena, G., Sabatowski, R.,
Grond, S., Lehmann, K.A., Cleeland, C.S. (1999). Validation of the
German version of the Brief Pain Inventory. J Pain Symptom Manage
18, 180187.
Smeeth, L., Ng, E.S. (2002). Intraclass correlation coefficients for cluster
randomized trials in primary care: Data from the MRC Trial of the
Assessment and Management of Older People in the Community.
Control Clin Trials 23, 409421.
Takai, Y., Yamamoto-Mitani, N., Okamoto, Y., Koyama, K., Honda, A.
(2010). Literature review of pain prevalence among older residents of
nursing homes. Pain Manag Nurs 11, 209223.
Tarzian, A.J., Hoffmann, D.E. (2004). Barriers to managing pain in the
nursing home: Findings from a statewide survey. J Am Med Dir Assoc
5, 8288.
The American Geriatrics Society (2002). The management of persistent
pain in older persons. J Am Geriatr Soc 50, S205S224.

2015 European Pain Federation - EFIC

Appropriate pain management in nursing home residents

The American Geriatrics Society (2009). Pharmacological management

of persistent pain in older persons. Pain Med 10, 10621083.
Wears, R.L. (2002). Advanced statistics: Statistical methods for
analyzing cluster and cluster-randomized data. Acad Emerg Med 9,
330341.
Won, A.B., Lapane, K.L., Vallow, S., Schein, J., Morris, J.N., Lipsitz,
L.A. (2004). Persistent nonmalignant pain and analgesic prescribing
patterns in elderly nursing home residents. J Am Geriatr Soc 52, 867
874.
Wood, B.M., Nicholas, M.K., Blyth, F., Asghari, A., Gibson, S. (2010).
Assessing pain in older people with persistent pain: The NRS is valid
but only provides part of the picture. J Pain 11, 12591266.
Wulff I., K
onner F., K
olzsch M., Budnick A., Drager D., Kreutz R.
(2012) [Interdisciplinary guidance for pain management in nursing
home residents]. Z Gerontol Geriatr, 45, 505544.

Web references
WHO Collaborating Centre for Drug Statistics Methodology, Oslo
Norway. Anatomic Therapeutic Chemical (ATC) classification index
with Defined Daily Doses (DDDs). http://www.whocc.no/
atc_ddd_index/ (20 June 2013).

Eur J Pain 19 (2015) 1331--1341

1341

Copyright of European Journal of Pain is the property of Wiley-Blackwell and its content may
not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's
express written permission. However, users may print, download, or email articles for
individual use.