Journal of Pain and Symptom Management
Volume 27, Issue 2 , Pages 104-113, February 2004

Pain characteristics and treatment outcome for advanced cancer patients during the first week of specialized palliative care

Department of Palliative Medicine, Bispebjerg Hospital, Copenhagen, Denmark

Accepted 30 June 2003.

Article Outline

Abstract 

To examine pain in cancer patients referred for specialized palliative care, we described pain characteristics and medication on admission, examined changes in pain during the first week, and searched for predictors of initial pain intensity and treatment outcome. On arrival in the department (T0) and after one week (T1), pain was evaluated with the Edmonton Symptom Assessment System (ESAS) and EORTC QLQ-C30. Analgesics were recorded. We investigated the associations between initial pain scores, as well as differences from T0 to T1, and clinical and sociodemographic parameters, initial medication, and medical interventions. Of 267 eligible patients, initial pain scores were obtained from 175. Initial pain scores were high, although 81% of patients received opioid treatment at T0. Bone metastases, neuropathic pain, mixed pain pathophysiology, and breakthrough pain were associated with higher initial pain scores. Pain scores decreased during the first week. No single parameter convincingly predicted a better or worse outcome of pain treatment.

Keywords:  Pain, pain pathophysiology, palliative care, advanced cancer, analgesic medication, self-assessment questionnaire, quality of life, treatment outcome

 

Back to Article Outline

1. Introduction 

In palliative care units, patients with advanced, incurable cancer present with a diversity of primary tumors, metastatic locations, previous treatments, and resulting symptomatology. Although many symptoms may be present in advanced cancer, pain is probably the most feared and, when present, often the most distressing symptom towards the end of life.

Epidemiological data on pain prevalence,1., 2., 3., 4. pain etiology and pathophysiology,5., 6., 7., 8. and pain treatment9., 10. in advanced cancer patients are increasingly available. Data on the pain treatment outcome in these patients are also available,11., 12., 13., 14., 15. although such data are only rarely correlated with clinical or sociodemographic parameters16 or (outside of clinical trials) with information on medical intervention.17

Denmark is known as a country with a liberal approach towards opioid use and has the highest opioid consumption per capita in the world.10., 18. Therefore, most patients with advanced, incurable cancer and pain are likely to be receiving opioid treatment before referral to a department of palliative medicine. This indicates that patients referred to specialized palliative care are patients with serious pain problems in spite of opioid treatment. To our knowledge, such a heavily afflicted patient population has not been investigated before regarding intensity and pathophysiology of pain, associations with clinical and sociodemographic factors, and treatment outcome.

In the present report, we examine pain in patients admitted to a department of palliative medicine by 1) giving a detailed description of pain characteristics and pain medication at admission, 2) examining the changes in pain intensity during the first week, and 3) exploring possible predictors of initial pain intensity and treatment outcome.

Back to Article Outline

2. Methods 

2.1. Patients 

Patients referred to our specialist palliative care unit between June 1998 and June 2000 were included in the study provided they were Danish-speaking, over 18 years of age, and able to give informed consent. The study was approved by the local Ethics Committee.

2.2. Assessments 

2.2.1. Clinical and sociodemographic data 

From all referred patients, clinical and sociodemographic parameters were registered and the doctor assessed the patient's Karnofsky Performance Status (KPS).19 We registered the length of survival from first encounter with the department.

2.2.2. Pain characteristics 

The doctor assessed the etiology and the pathophysiology of the pain. The doctor asked if the patient experienced breakthrough pain (yes/no) and to what extent (1=not at all, 2=a little, 3=quite a bit, 4=very much) the patient had experienced pain at rest, pain with movement, and interruption of sleep by pain. The patients were asked to name and prioritize up to five symptoms that they considered the most troublesome.

2.2.3. Patient self-assessment 

On the day of first contact with the department, the patients were asked if they would participate in the study. Consenting patients were given the self-assessment questionnaires Edmonton Symptom Assessment System (ESAS),20 the EORTC Quality of Life Questionnaire (EORTC QLQ-C30, version 3.0),21 and the Hospital Anxiety and Depression Scale (HADS),22 and were assessed by the doctor using the Mini Mental State Examination (MMSE).23 To facilitate patient participation even if they were willing or able to fill in relatively few items, the patients could choose whether they wished to complete all of the questionnaires or a subset of items only. In this article, a “participant” is defined as a consenting patient who completed at least one of the pain scales at the day of first contact or the next day.

The patients were assessed weekly for up to 13 weeks. In this article, the assessments from the day of first contact or the next day (days 0 or 1; T0) and one week after (days 7 or 8; T1) are reported. Questionnaires completed outside these time windows were excluded. The pain scores of EORTC QLQ-C30 and ESAS are reported as parallel outcome measures in this study.

The mean differences in pain scores from T0 to T1 (ΔEORTC and ΔESAS) were calculated for patients who completed questionnaires at both time points. We investigated the association between initial pain scores as well as the differences in pain scores and a number of clinical and sociodemographic parameters in order to reveal variables that would predict a higher initial score or a better response to pain treatment.

2.3. Medical intervention 

Drugs, doses, and administration routes of pain medication were registered weekly. As for drugs that can be used as adjuvant analgesics, we recorded the drugs and doses regardless of the main reason for prescription (pain, depression, epilepsy, etc.). This is because such drugs may have an analgesic effect regardless of the indication for prescription, and because the drugs were often prescribed for more than one reason.

The regularly scheduled analgesic doses could be recorded for almost all patients. As for the analgesic rescue doses, we recorded the administered rather than the prescribed doses. The administered rescue doses were registered in the nurses' records and could be reported only for inpatients and only at T1. For outpatients and patients in palliative home care, there was not sufficient information on the administered rescue doses. Opioid doses were converted to parenteral morphine equivalents. Changes in pain medication were registered according to “main intervention”: change in opioid dose, drug, administration route, and addition or discontinuation of adjuvant pain medication.

2.4. Statistics 

Non-parametric analyses were used to compare distributions in groups, as most scores were non-normally distributed. Data were analyzed using the SAS statistical package 8.02. Wilcoxon tests were used for the comparisons of pain scores in groups defined by sociodemographic (sex, age, referring departments, etc), clinical (KPS, cognitive and emotional function, survival, etc.), and pain-related (pain etiology and pathophysiology, pain prioritization, etc.) parameters.

In order to separate the effect of different factors from each other, multivariate regression analyses were performed (ordinal logistic regression for EORTC QLQ-C30 scores and linear regression for ESAS scores). For the initial pain scores, the analysis included the variables showing a significant difference in the univariate analyses for at least one of the initial pain scores (EORTC QLQ-C30 or ESAS). Similarly, the multivariate analyses for ΔEORTC and ΔESAS included the variables showing a significant difference in the univariate analyses for at least one of the two outcomes. To avoid colinearity between the initial pain scores and variables very closely linked to the degree of pain (pain prioritization, degree of pain at movement, pain at rest, as well as sleep disturbance due to pain), the multivariate regression analysis of the initial pain scores was repeated with these variables excluded, in spite of their statistical significance in the univariate analyses.

Back to Article Outline

3. Results 

Of the 278 patients referred, ten patients did not speak Danish and one patient was less than 18 years old. Sixty-six (25% of the 267 eligible patients) declined participation or were considered too ill to ask. Of 201 patients consenting to participate, 175 (66% of eligible patients) completed at least one of the two pain scales at T0 and were considered participants. One hundred forty-two (53%) completed the pain scale in the EORTC-QLQ-C30 and 167 (63%) the pain item in the ESAS. Sociodemographic and clinical data for the participants are shown in Table 1.

Table 1. Sociodemographic and Clinical Data for 175 Participating Patients

At T1, 8 participants had died, and 117 of the remaining 167 patients completed the EORTC QLQ-C30 pain scale and 121 the ESAS pain item. Ninety-eight patients completed the pain scale of the EORTC-QLQ-C30 and 116 the ESAS pain item at T0 as well as T1.

3.1. Initial pain characteristics 

3.1.1. Pain etiology 

According to the medical assessment, most patients (n=144) had pain caused by tumor growth and/or pressure. Of these, 29 patients also had other causes of pain: antineoplastic treatment (18), disease-related conditions (postherpetic neuralgia, decubitus, candidiasis, etc.) (8), or pain unrelated to the malignant disease (7).

3.1.2. Pain pathophysiology 

Sixty-two percent of our patients were diagnosed to have somatic pain, 45% neuropathic pain, and 28% visceral pain. Forty-one patients were judged to have pure somatic pain, 16 pure neuropathic pain, and 24 pure visceral pain. Seventy-two patients had mixed pain pathophysiology: 21 patients had somatic/visceral pain conditions, and of these, two patients had also neuropathic pain. Forty-seven patients had somatic/neuropathic pain conditions, and four had neuropathic/visceral pain conditions. Two patients were judged to have psychogenic pain, one combined with somatic pain, and one with visceral pain. For 22 patients, there was no information on pain pathophysiology or no pain.

3.1.3. Breakthrough pain 

From 121 patients, there was information of whether or not the patient had breakthrough pain. Of 73 patients (60%) with breakthrough pain, 41 patients had activity or movement as the main precipitating factor. For eight patients, inadequate opioid treatment and for six patients, constipation, hiccups, bladder tenesmi, or gout were the reasons given for the breakthrough pain. For 16 patients, no precipitating factor for breakthrough pain was stated.

Fifty-five (31%) patients had “quite a bit” or “very much” pain at rest, and 117 (67%) had “quite a bit” or “very much” pain at movement. For 38 patients (22%), sleep was interrupted “quite a bit” or “very much” by pain.

3.1.4. Importance of pain relative to other symptoms/problems 

Of the 175 patients, 77 patients (44%) selected pain as being their worst symptom, 47 (27%) as important but not worst (priority 2–5), and 51 (29%) as not among the 5 most troublesome symptoms.

3.2. Pain medication 

At T0, information on medication was available from 164 of the 175 participants. Grouping of patients according to the World Health Organization (WHO) analgesic ladder, as well as administration routes, is shown in Table 2. At T0, 81% of patients received regularly scheduled opioids, compared to 91% at T1. Daily doses of the different opioids used in our patients at T0 and T1 are shown in Table 3. In the same table, the daily doses are converted to equianalgesic doses of parenteral morphine.

Table 2. Pharmacological Pain Treatment in 175 Advanced Cancer Patients
On Arrival (n=164)After One Week (n=149)
Pain Treatmentn%n%
WHO-step
No analgesics127.385.4
Step 1 (weak analgesics)116.753.4
Step 2 (weak opioids)84.910.7
Step 3 (strong opioids)13381.113691.3
Adjuvant analgesics
Antidepressants4929.94630.9
Anticonvulsants2213.42516.8
NMDA antagonists (ketamine)0042.7
Administration routes for opioids
Oral9658.58355.7
Rectal42.421.3
Transdermal3622.03422.8
Parenteral21.2149.4
Spinal/epidural42.4106.7
Table 3. Opioid Consumption in 175 Patients with Advanced Cancer
On First Contact, n=164 Dose (mg/day) After One Week, n=149 Dose (mg/day)
EDMnMeanMedianRangenMeanMedianRange
Morphine
IR oral/rectal20/201030.9254–903549.0305–220
SR oral2077144.812010–60063134.88020–600
Parenteral10120.020 1884.3602.5–340
Epidural/spinal1453.4153.75–1807120.36210–480
Methadone
Oral208144.752.537.5–3508170.982.515–400
Parenteral100 12525
Ketobemidon
IR oral/rectal15/10351.74015–110446.35020–65
SR oral151110.0110 260.06030–90
Parenteral7.52237.5 100–3750
Fentanyl
Transdermal0.1362.72.40.6–7.2342.62.40.6–7.2
Parenteral0.10 31.51.00.05–3.5
Oxycodone
IR oral1515.05 15.05
SR oral15235.0 20–501180.96010–180
Sufentanil
Parenteral0.010 70.40.30.1–1.0
Daily equianalgesic doses of parenteral morphine 133141.0753.3–1800136213.2907.5–4800

EDM=Equipotent dose of opioid corresponding to morphine 10 mg SC/IM;18., 33., 34., 35. IR=immediate release; SR=sustained release.

3.2.1. Types of intervention 

In 51 patients, the opioid dose was increased; 28 patients underwent opioid rotation, and in four patients, the route of opioid administration was changed. In 40 patients, the opioid dose was unchanged (<30% change) and in 10 patients, the dose was reduced. Antidepressants were added in 14 patients and anticonvulsant drugs added in 17 patients. The antidepressants most frequently used at T0 were amitriptyline (18 patients), nortriptyline (16 patients), and citalopram (18 patients), and these figures were largely unchanged at T1. Valproate was the anticonvulsant most frequently used (16 patients at T0 and 21 patients at T1), and other anticonvulsants were used by a maximum of two patients.

3.3. Changes in pain during the first week 

Table 4 (upper row) shows high initial pain scores according to both questionnaires. There was a statistically significant decrease in the mean pain scores of EORTC QLQ-C30 (P<0.0001) and ESAS (P=0.0013) from arrival in the department to one week later (Table 4). According to the EORTC QLQ-C30, 74% had pain scores ≥60 at T0, compared to 55% at T1. The corresponding figures for the ESAS were 40% at T0 and 34% at T1.

Table 4. Self-Assessed Pain Scores and Difference in Pain Scores (score T1−score T0) for All 175 Participants and for Participants Subdivided by Sociodemographic and Clinical Variables (at T0)
EORTC QLQ-C30ΔEORTC QLQ-C30ESASΔESAS
NcMeanNcMeanNcMeanNcMean
All participants1427098−1216750116−9
Men657044−19a734956−11
Women777054−6945160−7
Age
<606276a39−157755a52−13
≥60806659−10904664−5
MMSE
≤23347121−27b394526−10
>23976970−81195085−7
HAD depression
0–74164a31−11395029−15
8–10306623−10264621−12
11–21627739−13615039−4
Pain
worst symptom6585b46−137664b56−13a
prioritized, not worst357124−14445430−12
not prioritized424728−84725304
Pain etiology
Tumor11675b79−1413756b96−11b
Other264919−43024204
Bone metastases
Present4678a35−165659b48−14a
No bone metastases966663−101114668−5
Breakthrough pain
Present5782b44−166960a53−15
No breakthrough pain406426−8464831−8
Pain qualities
01744b13−42123b141
1656841−14754750−7
2587943−13696151−14
3283117255120
Neuropathic pain
Present5483b39−156663b51−14a
No neuropathic pain886259−101014165−5
Pain at rest
None2847b19−103525b2211b
A little667345−15755358−15
Quite a bit338024−10386123−10
Very much11828−10156911−12
Pain at movement
None1635b12−71812b129a
A little286119−12303521−9
Quite a bit517632−14585832−10
Very much438533−14566348−13
Sleep disturbed by pain
None7162b48−117738b53−3a
A little disturbed367826−11495734−10
Quite a bit227716−16245817−15
Very much10936−25148310−25
Medical interventions
Opioid rotation2388b16−162765b21−15
No rotation1106575−111294688−7
Anticonvulsant
Prescribed1286a10−81571b13−19
Not prescribed1216881−131414796−7
WHO analgesic ladder at T0
No analgesics1139b8−151123b7−1
Step 17335310326−13
Step 28637−96395−2
Step 31077571−131295491−9
MED at T0
0 mg3554b27−93839b25−8
1–50 mg437028−15484834−11
51–150 mg387825−10465632−10
>150 mg267918−14355725−5

ΔEORTC QLQ-C30 and ΔESAS=score at T1−score at T0.

Nc: Number of patients from a given group completing the pain scale/item; MMSE: Mini Mental State Examination; HADS: Hospital Anxiety and Depression Scale; MED: Morphine Equivalent Dose. P values for test for difference between groups:

aP<0.05, and bP<0.01 in univariate analyses.

3.4. Possible predictors of pain characteristics and treatment outcome 

According to the univariate analyses, younger patients, patients who described pain as one of the most troublesome symptoms, patients with mixed pain pathophysiology or more causes of pain, and patients with neuropathic pain, bone metastases, or breakthrough pain had significantly higher initial pain scores on both questionnaires (Table 4). Patients who had pain because of tumor growth and/or pressure had higher pain scores than those who had other pain etiologies. The presence/absence of pain pathologies other than neuropathic pain (somatic, visceral, or psychogenic) was not associated with significant differences in pain intensity levels. Patients who underwent opioid rotation or who had an anticonvulsant drug prescribed had higher initial pain scores. Patients who were receiving strong opioid treatment on arrival in the department (step 3 on the WHO analgesic ladder) had higher initial pain scores than the patients on steps 1 or 2, and the patients who were treated with the highest opioid doses had the highest pain scores. There seemed to be no association between initial pain levels and sex, KPS, vocational training, cohabitation, department of referral, length of survival, or HADS anxiety scores.

The first multivariate regression analysis revealed three variables to significantly predict the initial pain score measured by the EORTC QLQ-C30: HADS depression score (P=0.049), pain prioritization at T0 (P<0.0001), and degree of pain at movement (P=0.028). In the multivariate model, the only variable associated with initial ESAS pain score was pain prioritization at T0 (P<0.0001). Two of these predictors were considered closely linked with the dependent variables. We therefore performed a second multivariate analysis excluding these and two other independent variables (degree of sleep disturbance due to pain and pain at rest). This second multivariate analysis revealed the following results: For the EORTC QLQ-C30 initial pain scores, neuropathic pain (P=0.007), breakthrough pain (P=0.030), and morphine equivalent dose at T0 (P=0.026) were predictive. For the ESAS pain scores, only presence of neuropathic pain was predictive (P=0.007).

Few variables were convincingly associated with the degree of pain relief. There was a tendency towards men having better pain relief than women (significant for ΔEORTC only) and patients with tumor-related pain, as well as patients with neuropathic pain, having better pain relief (only ΔESAS significant) (Table 4). We found no variables concerning the analgesic treatment to predict better or poorer pain relief. Multivariate regression analysis revealed two variables to significantly predict the ΔEORTC: sex (P=0.0047), and cognitive functioning (P=0.009). For the ΔESAS, only the degree of pain at rest was predictive (P=0.0043).

Back to Article Outline

4. Discussion 

4.1. Initial pain characteristics 

The intensity of pain found in participating patients was high. The mean EORTC QLQ-C30 pain scale score was 70. Mystakidou et al.11 found a mean EORTC QLQ-C30 pain score of 88 in cancer patients referred to an outpatient clinic in a palliative care unit. Quite differently, Jordhoy et al.24 found a mean EORTC QLQ-C30 pain score of 48 in 395 patients with advanced cancer. A possible explanation for these differences is the criterion for admission to Jordhoy's study, which were less strict than described in our study (patients were eligible if they had incurable cancer and had a life expectancy between 2 and 9 months, whereas we required “pronounced symptomatology” and had no criteria as to life expectancy).

The distribution in pain pathophysiology in our study was somewhat different from the findings of another study, which had fewer patients with neuropathic pain and more patients with somatic or visceral pain.6 The distribution of pain etiology in our study is in accordance with the findings of other studies.5., 6., 8.

The prevalence of breakthrough pain was 60%. Other researchers have found prevalences ranging from 39% to 93%.8., 25., 26., 27., 28., 29. The most frequently stated precipitating factor for breakthrough pain was activity/movement (56%). This is comparable with the findings of other studies, in which activity/movement precipitated 20–54% of breakthrough pain episodes.25., 26., 28.

4.2. Pain medication 

These patients had access to opioids (Table 2) but still had significant pain (Table 4). Hence opioid availability is just one of the necessary factors required to provide good pain control. Mercadante recorded the use of analgesics in 2,500 advanced cancer patients referred to a palliative home care service.30 At referral, only 9.8% of the patients were treated with opioids, compared to 35.2% in the last week of life. According to a recent study by Meuser et al.,14 25.3% of advanced cancer patients referred to a pain clinic were treated with opioids, compared to 52.6% in the middle of the treatment period (mean 24.5 days later).

The frequency of using adjuvant analgesics is higher in our study than in the study by Meuser et al.14 but was less different from Mercadante's results.30

4.3. Pain relief during the first week 

In the study by Mystakidou et al.,11 exceptional decreases in symptomatology were found during the first two weeks of treatment in a palliative care department in Greece: mean pain levels decreased from 88 to 38 two weeks later. Prior pain medication and interventions were not described. Meuser et al. reported pain scores decreasing from a mean of 65.6 (0–100) to 20.9 in a similar time interval.14 Due to the differences in case mix between ours and other published studies, we cannot know whether our treatment could be improved to match that given in other centers or whether the relatively modest reduction in pain scores seen in our study (Table 4) is mainly a reflection of the selection of highly treatment resistant patients. Results from other institutions treating similar groups of patients are needed to answer this question. The latter interpretation, however, is corroborated by the finding that most of our patients were in the poor prognosis group (see below).

4.4. Possible predictors for pain characteristics and relief 

4.4.1. Baseline pain scores 

Several parameters seemed to be associated with high pain scores in the univariate analyses. Many of these variables are closely linked to the each other and also to the patients' pain scores, and the first multivariate analysis disclosed that the pain prioritization at T0 (worst symptom, among the five worst symptoms but not worst, or not prioritized) was the main predictor of pain score. This seems to be a rather predictable finding: the higher the pain prioritization, the higher pain score, and vice versa. The same can be said for the other predictor (significant for ΔEORTC only), pain at movement. Therefore, as mentioned above, a second multivariate analysis was performed, excluding the independent variables with most suspected colinearity with the dependent variables. This second multivariate analysis revealed one variable to be a significant predictor of initial high or low pain score in both questionnaires: presence or absence of neuropathic pain. In a survey of cancer patients with pain, Caraceni et al.8 found that, in univariate analyses, breakthrough pain, somatic pain or neuropathic pain, young age, and low performance status were associated with higher pain scores. In multivariate analyses, however, only breakthrough pain, somatic pain, and low performance status predicted intense pain.

4.4.2. Treatment outcome 

It is notable that, as shown in Table 4, we found no single factor for which the presence or absence could determine the extent of pain relief measured on both questionnaires. The change in EORTC pain score was significantly different in men and women, suggesting that women did not achieve very effective pain relief. The same tendency was present but not significant for ΔESAS. Patients with neuropathic pain had significantly better relief according to ESAS but not EORTC QLQ-C30. There has been a general belief that patients with neuropathic pain are less responsive to opioid treatment,16 but some researchers have shown that patients with neuropathic pain do not have any particular disadvantage compared to patients with somatic or visceral pain conditions.15., 31., 32.

The multivariate analysis of ΔEORTC indicated that men and patients with a poor cognitive function would have a better treatment outcome, whereas the analysis of ΔESAS indicated that patients with much sleep disturbance due to pain would be those who experienced the best treatment effect. We must conclude that in the present patient material, univariate and multivariate regression analyses revealed no variables that convincingly predicted a better or poorer treatment outcome.

According to Bruera et al.,16 patients with neuropathic pain, breakthrough pain, psychological distress (anxiety or depression), fast opioid dose escalation, or addiction to alcohol/drugs, would have a poorer prognosis as for pain relief. We had no general information on addictive personality or on changes in opioid dose up to admission in the department. However, according to presence of neuropathic pain, breakthrough pain, or psychological distress (defined as score 8–21 on the anxiety or depression scales of the HADS, corresponding to possible or definite anxiety or depression), as many as 154 (88% of participants) were in the “poor prognosis” group. This might also explain the more modest treatment results in our study, compared to other studies. The prognosis categorization (according to responsiveness to NSAID, and in case of unresponsiveness, to opioids) suggested by Mercadante et al.17 seems inapplicable to our patient material: most of our participants (81%) would be excluded as they were in opioid treatment on arrival in the department.

4.4.3. Two parallel measures of pain intensity 

In the present study we have presented two parallel pain endpoints, the ESAS and the EORTC pain scales. There were minor differences in the results obtained with the two measures but no knowledge is available as to which of the two measures is most valid. Therefore, even though our data presentation and conclusions would have been more straightforward if only one endpoint had been reported, we chose to report them in parallel. This yields the advantage that the results can be compared with new studies using either measure. Furthermore, investigation of predictors of pain includes multiple hypothesis testing. Taking the conservative point of view of trusting only findings made on both measures reduces the number of false positive results.

Back to Article Outline

5. Conclusion 

We have described a group of advanced cancer patients with high initial pain intensity despite liberal access to opioids in Denmark. Univariate analyses showed that the presence of bone metastases, neuropathic pain, more pain etiologies, mixed pain pathophysiology, or breakthrough pain were associated with higher initial mean pain scores. Multivariate analyses revealed neuropathic pain to be the only variable to predict the level of initial pain scores on both measures of pain. No clinical, sociodemographic, or therapeutic variables could predict the extent of pain relief seen on both measures of pain (EORTC QLQ-C30 and ESAS). Despite recent advances in multidisciplinary pain treatment, there is room for improvement of the treatment of the patients with poorest pain prognosis: further research focused on this group of patients is needed and should be coupled with careful monitoring of the pain characteristics and treatment outcome of consecutive patients in as many institutions as possible.

Back to Article Outline

Acknowledgements 

This study was supported by a grant from the Danish Cancer Society (no. 98 150 07) and the research fund of Copenhagen Hospital Corporation.

Back to Article Outline

References 

  1. Walsh D, Donnelly S, Rybicki L. The symptoms of advanced cancer: relationship to age, gender, and performance status in 1,000 patients. Support Care Cancer. 2000;8:175–179
  2. Vainio A, Auvinen A. Prevalence of symptoms among patients with advanced cancer: an international collaborative study. Symptom Prevalence Group. J Pain Symptom Manage. 1996;12:3–10
  3. Daut RL, Cleeland CS. The prevalence and severity of pain in cancer. Cancer. 1982;50:1913–1918
  4. Schuit KW, Sleijfer DT, Meijler WJ, et al.  Symptoms and functional status of patients with disseminated cancer visiting outpatient departments. J Pain Symptom Manage. 1998;16:290–297
  5. Banning A, Sjogren P, Henriksen H. Pain causes in 200 patients referred to a multidisciplinary cancer pain clinic. Pain. 1991;45:45–48
  6. Grond S, Zech D, Diefenbach C, et al.  Assessment of cancer pain: a prospective evaluation in 2266 cancer patients referred to a pain service. Pain. 1996;64:107–114
  7. Ripamonti C, Zecca E, Brunelli C, et al.  Pain experienced by patients hospitalized at the National Cancer Institute of Milan: research project “towards a pain-free hospital.”. Tumori. 2000;86:412–418
  8. Caraceni A, Portenoy RK. An international survey of cancer pain characteristics and syndromes. IASP Task Force on Cancer Pain. International Association for the Study of Pain. Pain. 1999;82:263–274
  9. Portenoy RK, Lesage P. Management of cancer pain. Lancet. 1999;353:1695–1700
  10. Sjogren P, Banning AM, Jensen NH, et al.  Management of cancer pain in Denmark: a nationwide questionnaire survey. Pain. 1996;64:519–525
  11. Mystakidou K, Tsilika E, Parpa E, et al.  The EORTC core quality of life questionnaire (QLQ-C30, version 3.0) in terminally ill cancer patients under palliative care: validity and reliability in a Hellenic sample. Int J Cancer. 2001;94:135–139
  12. Chiu TY, Hu WY, Chen CY. Prevalence and severity of symptoms in terminal cancer patients: a study in Taiwan. Support Care Cancer. 2000;8:311–313
  13. Lo RS, Ding A, Chung TK, et al.  Prospective study of symptom control in 133 cases of palliative care inpatients in Shatin Hospital. Palliat Med. 1999;13:335–340
  14. Meuser T, Pietruck C, Radbruch L, et al.  Symptoms during cancer pain treatment following WHO-guidelines: a longitudinal follow-up study of symptom prevalence, severity and etiology. Pain. 2001;93:247–257
  15. Banning A, Sjogren P, Henriksen H. Treatment outcome in a multidisciplinary cancer pain clinic. Pain. 1991;47:129–134
  16. Bruera E, Schoeller T, Wenk R, et al.  A prospective multicenter assessment of the Edmonton staging system for cancer pain. J Pain Symptom Manage. 1995;10:348–355
  17. Mercadante S, Maddaloni S, Roccella S, et al.  Predictive factors in advanced cancer pain treated only by analgesics. Pain. 1992;50:151–155
  18. Clausen TG, Eriksen J, Borgbjerg FM. Legal opioid consumption in Denmark 1981–1993. Eur J Clin Pharmacol. 1995;48:321–325
  19. Yates JW, Chalmer B, McKegney FP. Evaluation of patients with advanced cancer using the Karnofsky performance status. Cancer. 1980;45:2220–2224
  20. Bruera E, Kuehn N, Miller MJ, et al.  The Edmonton Symptom Assessment System (ESAS): a simple method for the assessment of palliative care patients. J Palliat Care. 1991;7:6–9
  21. Aaronson NK, Ahmedzai S, Bergman B, et al.  The European Organization for Research and Treatment of Cancer QLQ-C30: a quality-of-life instrument for use in international clinical trials in oncology. J Natl Cancer Inst. 1993;85:365–376
  22. Zigmond AS, Snaith RP. The hospital anxiety and depression scale. Acta Psychiatr Scand. 1983;67:361–370
  23. Folstein MF, Folstein SE, McHugh PR, “Mini-mental state” . A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975;12:189–198
  24. Jordhoy MS, Fayers P, Loge JH, et al.  Quality of life in advanced cancer patients: the impact of sociodemographic and medical characteristics. Br J Cancer. 2001;85:1478–1485
  25. Petzke F, Radbruch L, Zech D, et al.  Temporal presentation of chronic cancer pain: transitory pains on admission to a multidisciplinary pain clinic. J Pain Symptom Manage. 1999;17:391–401
  26. Portenoy RK, Hagen NA. Breakthrough pain: definition, prevalence and characteristics. Pain. 1990;41:273–281
  27. Portenoy RK, Payne D, Jacobsen P. Breakthrough pain: characteristics and impact in patients with cancer pain. Pain. 1999;81:129–134
  28. Swanwick M, Haworth M, Lennard RF. The prevalence of episodic pain in cancer: a survey of hospice patients on admission. Palliat Med. 2001;15:9–18
  29. Fine PG, Busch MA. Characterization of breakthrough pain by hospice patients and their caregivers. J Pain Symptom Manage. 1998;16:179–183
  30. Mercadante S. Pain treatment and outcomes for patients with advanced cancer who receive follow-up care at home. Cancer. 1999;85:1849–1858
  31. Mercadante S. Predictive factors and opioid responsiveness in cancer pain. Eur J Cancer. 1998;34:627–631
  32. Jadad AR, Carroll D, Glynn CJ, et al.  Morphine responsiveness of chronic pain: double-blind randomized crossover study with patient-controlled analgesia. Lancet. 1992;339:1367–1371
  33. Coda BA, O'Sullivan B, Donaldson G, et al.  Comparative efficacy of patient-controlled administration of morphine, hydromorphone, or sufentanil for the treatment of oral mucositis pain following bone marrow transplantation. Pain. 1997;72:333–346
  34. Hanks G, Cherny N. Opioid analgesic therapy. In:  Doyle D,  Hanks G,  Macdonald N editor. Oxford textbook of palliative medicine. Oxford: Oxford University Press; 1998;p. 331–355
  35. Pereira J, Lawlor P, Vigano A, et al.  Equianalgesic dose ratios for opioids. a critical review and proposals for long-term dosing. J Pain Symptom Manage. 2001;22:672–687

PII: S0885-3924(03)00513-X

doi:10.1016/j.jpainsymman.2003.06.011

Journal of Pain and Symptom Management
Volume 27, Issue 2 , Pages 104-113, February 2004

Article Tools