Prevalence and Characteristics of Breakthrough Pain in Cancer Patients Admitted to a Hospice

  • Published on

  • View

  • Download

Embed Size (px)


<ul><li><p> U.S. Cancer Pain Relief Committee, 2000 0885-3924/00/$see front matterPublished by Elsevier, New York, New York PII S0885-3924(00)00161-5</p><p>Vol. 20 No. 2 August 2000 Journal of Pain and Symptom Management 87</p><p>Original Article</p><p>Prevalence and Characteristics of Breakthrough Pain in Cancer Patients Admitted to a Hospice</p><p>Giovambattista Zeppetella, BSc (Hons), MRCGP,Catherine A. ODoherty, PhD, MRCP (UK), and Silke Collins, MD</p><p>Palliative Medicine Service (G.Z., C.A.D., S.C.), St Josephs Hospice; and Royal Hospitals NHS Trust (G.Z.), London, United Kingdom</p><p>Abstract</p><p>A prospective survey was undertaken to determine the prevalence and characteristics of breakthrough pain in cancer patients admitted to a hospice. Of 414 consecutive admissions, 33 patients were confused or too unwell to take part and 136 were pain-free. The remaining 245 reported 404 pains (range 15 per patient); of these patients, 218 (89%) had breakthrough pain and identified 361 pains (range 15 per patient). Breakthrough pain was classified as somatic (46%) visceral (30%), neuropathic (10%) or mixed etiology (16%). Thirty-eight percent of pains were severe or excruciating. The average number of daily breakthrough pain episodes was 7 (range 114); 49% occurred suddenly. Most (59%) were unpredictable, and 72% lasted less than 30 minutes. Seventy-five percent of patients were dissatisfied with their pain control. Breakthrough pain is common among patients admitted to our hospice. It is frequent, short lasting, often unpredictable and not necessarily related to chronic pain making treatment difficult. </p><p>J Pain Symptom Manage 2000;20:8792.</p><p> U.S. Cancer Pain Relief Committee, 2000.</p><p>Key Words</p><p>Cancer pain, breakthrough pain, hospice</p><p>Introduction</p><p>Pain is one of the most common and fearedsymptoms of cancer. Many patients have morethan one type of pain; not all pains are causedby the cancer. A number of studies have estab-lished that pain can be controlled in the major-ity of patients using the World Health Organi-zation guidelines.</p><p>14</p><p> Cancer patients can,however, experience breakthrough pain de-spite adequate background analgesia.</p><p>Breakthrough pain is a transitory exacerba-tion in otherwise acceptable analgesia for stablepain.</p><p>5</p><p> It can be difficult to manage and can com-promise good pain control.</p><p>6</p><p> Previous studieshave characterized breakthrough pain in pa-tients attending a cancer center,</p><p>7,8</p><p> a pain clinic,</p><p>9</p><p>and in hospice home care patients; </p><p>10</p><p> prevalencevaries between 4086%. The aim of this studywas to examine prospectively the prevalence andcharacteristics of breakthrough pains in termi-nally ill cancer patients admitted to a hospice.</p><p>Methods</p><p>A prospective survey of hospice admissionswas undertaken between February and August</p><p>Address reprint requests to: </p><p>Giovambattista Zeppetella,BSc (Hons), MRCGP, St. Josephs Hospice, MareStreet, London E8 4SA, United Kingdom.</p><p>Accepted for publication: September 29, 1999.</p></li><li><p> 88 Zeppetella et al. Vol. 20 No. 2 August 2000</p><p>1998; patients had an estimated prognosis ofsix months or less. All patients were assessed bythe authors using an agreed schedule adaptedfrom Portenoy and Hagen.</p><p>7</p><p> Patients were ex-cluded from the survey if they were confusedor too unwell to answer the questions or if theywere pain free. Approval was obtained fromthe local Ethics and Research Committee.</p><p>At assessment, patients were asked to de-scribe their chronic pains with respect to loca-tion, duration, and severity. Patients were thenasked to characterize their breakthroughpainif presentwith respect to location, se-verity, temporal relations, relationship to fixedanalgesic dose, precipitant, predictability, andpalliative factors. Patients were asked to basetheir answers on the previous 24-hour periodof time. A note was made of the likely causeand pathophysiology (i.e. somatic, visceral,neuropathic or mixed; this was inferred fromthe history of the pain in the context of the his-tory of the disease</p><p>11</p><p>) for both chronic andbreakthrough pain. A record was made of allthe prescribed medication and patients wereasked if they were satisfied with the manage-ment of their pain.</p><p>Results</p><p>During the study period, there were 432 ad-missions to the hospice, of which 414 (96%)were patients with cancer (Table 1). The mostcommon diagnoses were cancers of lung(27%), breast (9%), and prostate (9%), andcancer of unknown primary (9%). Of the 414cancer patients, 33 (8%) were unable to an-swer questions because of dementia (4), confu-sion (14), or final stage of the disease (15).Twenty-five of the 33 patients who were unable</p><p>to self-report had been prescribed oral orparenteral opioids prior to admission. A fur-ther 136 patients did not report chronic painon admission; of these 15% had been pre-scribed weak opioids and 64% had been pre-scribed strong opioids.</p><p>Two hundred and forty-five patients re-ported a total of 404 chronic pains of variableduration (Table 2); pains had persisted fordays (10%), weeks (55%), or months (35%).Of these patients, 218 (89%) reported break-through pain; 361 separate pains were identi-fied. Most chronic and breakthrough painswere tumor-related; 11% of breakthroughpains were unrelated to the reported chronicpains. Thirty-eight percent of the pains were ei-ther severe or excruciating.</p><p>The most common sites for breakthroughpain were abdomen (30%), lower limb (19%),back (16%), and chest (12%). The average</p><p>Table 1</p><p>Characteristics of Patients Surveyed</p><p>Males Females Total</p><p>All admissions 196 236 432Non-cancer patients 11 7 18Cancer patients 185 229 414Average age 71 71 71Range 3397 36100 33100Confused patients 10 23 33Number of cancer patients surveyed 175 206 381Pain-free patients 62 74 136Patients with chronic pain 113 132 245Patients with chronic pain and BTP 100 118 218</p><p>BTP </p><p>5</p><p> Breakthrough pain.</p><p>Table 2</p><p>Characteristics Shared by Chronic and Breakthrough Pain</p><p>Chronic Breakthrough</p><p>Number of pains identified 404 361Range per patient 15 15Inferred pathophysiology (%</p><p>a</p><p>)Somatic 199 (49) 163 (45)Visceral 109 (27) 106 (29)Neuropathic 26 (6) 34 (9)Mixed 70 (17) 58 (16)</p><p>Intensity (%)Slight 73 (18) 58 (16)Moderate 209 (52) 167 (46)Severe 117 (29) 128 (36)Excruciating 5 (1) 8 (2)</p><p>Etiology (%)Tumor-related 270 (67) 255 (71)Treatment-related 31 (8) 39 (11)Unrelated to tumor/treatment 103 (26) 67 (19)</p><p>a</p><p>Percentages rounded to whole numbers.</p></li><li><p> Vol. 20 No. 2 August 2000 Breakthrough Pain in Hospice Admissions 89</p><p>number of daily episodes of individual painswas seven (range 114) (Figure 1). Approxi-mately half the pains occurred suddenly. Seven-teen percent resulted from end-of-dose failureof scheduled analgesics. Most pains (59%) wereunpredictable. Approximately 73% of painslasted 30 minutes or less (Figure 2); neuro-pathic pains were particularly brief, with 91%lasting 30 minutes or less, compared with somaticand visceral pains (69% and 62%, respectively,with durations of 30 minutes or less).</p><p>Patients were asked what best relieved theirbreakthrough pain. The most common factorswere analgesics (57%) and lying still (32%), in61 pains these two factors were given together.In 14% of pains, no palliative factors wereidentified.</p><p>Of the 218 patients with breakthrough pain,34% were prescribed nonopioid analgesics,23% weak opioids, and 64% strong opioids.Of the patients receiving strong opioids, 99took long-acting opioids (slow release mor-phine or transdermal fentanyl) or diamor-phine infusions; of these, 43% were not pre-scribed rescue medication. Of the 54 patientsprescribed rescue medication, morphine wasthe most common (91%), followed by diamor-phine (6%) and phenazocine (2%). There</p><p>were 77 prescriptions for adjuvant analgesics;the most common were corticosteroids (35 pa-tients), antidepressants (19), antispasmodics(12), and anticonvulsants (7). These analgesicswere usually prescribed with other analgesics,except in 10 patients when they were pre-scribed independently. Eighteen patients (8%)with chronic pain were not prescribed any an-algesic preparation.</p><p>Patients were asked if they were satisfied withtheir pain control. Seventy-eight percent of pa-tients with chronic pain without breakthroughpain were satisfied, compared with 25% of pa-tients with breakthrough pain (</p><p>x</p><p>2</p><p>5</p><p> 31, df </p><p>5</p><p>1,</p><p>P</p><p>,</p><p> 0.0005). The characteristics for break-through pain in both satisfied and dissatisfiedpatients are shown in Table 3; there are no sig-nificant differences between the characteristicsof the pain, although dissatisfied patients havesignificantly more occurrences of break-through pains (</p><p>x</p><p>2</p><p>5</p><p> 31, df </p><p>5</p><p>4, </p><p>P</p><p>,</p><p> 0.0001).</p><p>Discussion</p><p>The term breakthrough pain is usually ap-plied to patients who are receiving scheduledopioid regimens to treat continuous pain.Some cancer-related pains however, for exam-</p><p>Fig. 1. Frequency of breakthrough pains in hospice patients on admission.</p></li><li><p> 90 Zeppetella et al. Vol. 20 No. 2 August 2000</p><p>ple, neuropathic pain, do not respond well toopioids and adjuvant analgesics, such as antide-pressants and anticonvulsants, may be more ef-fective.</p><p>12,13</p><p> For the purpose of this study, break-through pain was taken to be a transitoryexacerbation of pain experienced by the pa-tient who has relatively stable and adequatelycontrolled baseline pain.</p><p>5</p><p> Several types ofbreakthrough pain exist, including incidentpain and end-of-dose pain, both of which werereported by patients surveyed.</p><p>Eight percent of cancer patients admitted tothe hospice were unable to take part in the sur-vey because they were too unwell. Most pa-tients had been prescribed opioids and somewere likely to have experienced breakthroughpain. These patients require careful monitor-ing to identify breakthrough pain; clinicalsigns may be subtle and difficult to identify,even for experienced hospice staff.</p><p>Pain in cancer is often seen as inevitable. Inthis survey, 136 patients were pain-free; ofthese, 79% were taking opioids and, therefore,presumably had had pain at some time prior toadmission. Of the 245 patients with pain, 89%had breakthrough pain; this figure is higherthan in previous studies.</p><p>710</p><p> Studies vary in howbreakthrough pain is recorded; one study, for</p><p>example, included only pain of greater thanmoderate intensity.</p><p>7</p><p> Of the 225 patients re-porting mild or moderate pain in this survey,78% were dissatisfied with their pain control,suggesting that these pain intensities also needconsideration. Perhaps an appropriate work-ing definition is that breakthrough pain, likechronic pain, is whatever the patient says it isand occurs whenever the patient says it does.</p><p>Most breakthrough pains are frequent andshort-lasting. This was seen particularly in neu-ropathic pains and has implications for success-ful management. Ideally, medication should beabsorbed quickly and produce a rapid onset ofanalgesia with minimal adverse events. For inpa-tients, the analgesia needs to be instantly avail-able; this may not be possible as controlleddrugs are usually locked away.</p><p>Breakthrough pain is usually managed withoral or parenteral rescue medication given inaddition to regular scheduled analgesics.</p><p>1416</p><p>The only controlled trials of medication forbreakthrough pain are limited to oral transmu-cosal fentanyl citrate (OTFC),</p><p>17 </p><p>which is nowavailable in the United States but not in theUnited Kingdom. The fact that over 40% of pa-tients with long-acting opioids were not pre-scribed rescue medication suggests a need for</p><p>Fig. 2. Duration of breakthrough pains in hospice patients on admission.</p></li><li><p> Vol. 20 No. 2 August 2000 Breakthrough Pain in Hospice Admissions 91</p><p>continued education. In those patients pre-scribed rescue medication, short-acting mor-phine was the most common. However, break-through pain in most patients lasted less than30 minutes. Other patient groups have shownsimilar results: cancer center (median duration30 minutes</p><p>7</p><p>), pain clinic (mean duration 32minutes</p><p>9</p><p>) and hospice home care (mean dura-tion 52 minutes</p><p>10</p><p>). The rapid onset and shortduration of breakthrough pains, therefore,may not be appropriately treated with short-acting oral morphine, which can take up to anhour to produce analgesia and then lasts forfour hours after administration.</p><p>Both the presence and the number of break-through pains were related to dissatisfactionwith treatment in the patients surveyed; similarfindings have been noted elsewhere.</p><p>6</p><p> This un-derlines the importance of specifically includ-ing an assessment of breakthrough pain in theoverall assessment and management plan.</p><p>This study adds to the literature on break-through pain and shows that, as in other pa-</p><p>tient groups, it is a common problem amongterminally ill cancer patients admitted to ourhospice. Breakthrough pain occurs frequently,is short-lasting, is often unpredictable, and isnot necessarily related to chronic pain. Fur-thermore, patients with far-advanced diseasemay be unable to give an accurate history butcould still experience breakthrough pain. Allthese factors make the management of break-through pain in this patient group difficult.</p><p>Acknowledgments</p><p>We acknowledge the help and advice of theJanssen Research Foundation, particularlyMark Travers and Liz Wager, in the prepara-tion of this manuscript.</p><p>References</p><p>1. Takeda F. Results of field-testing in Japan of theWHO draft interim guidelines on relief of cancerpain. Pain Clin 1986;1:8389.</p><p>Table 3</p><p>Characteristics of Breakthrough Pain According to Satisfaction with Treatment</p><p>Satisfied Dissatisfied Total</p><p>Number of patients (% male) 55 (53) 163 (55) 218 (54)Number of pains (% total) 69 (19) 292 (81) 361Average (range) 1.3 (13) 1.8 (15) 1.7 (15)Intensity</p><p>Slight 14 44 58Moderate 35 132 167Severe 19 109 128Excruciating 1 7 8</p><p>OnsetSudden 31 146 177Gradual 38 146 184</p><p>Relationship to analgesic doseRelated 9 51 60Unrelated 60 301 301</p><p>PrecipitantNone 40 169 209Incident</p><p>a</p><p>23 91 114Non-volitional</p><p>b</p><p>6 32 38Predictability</p><p>Predictable 29 119 148Unpredictable 40 173 213</p><p>Inferred pathophysiologySomatic 35 128 163Visceral 20 86 106Neuropathic 4 30 34Mixed 10 48 58</p><p>EtiologyTumor-related 44 211 255Treatment-related 8 31 39Unrelated to tumor/treatment 17 50 67</p><p>a</p><p>Pain induced by patients voluntary actions (e.g. movement, swallowing).</p><p>b</p><p>Pain induced by involuntary events (e.g. temperature, bowel distension).</p></li><li><p> 92 Zeppetella et al. Vol. 20 No. 2 August 2000</p><p>2. Ventafridda V, Tamburini M, Caraceni C, De-Conno F, Naldi F. A validation study of the WHOmethod for cancer pain relief. Cancer 1987;59:850856.</p><p>3. Schug SA, Zech D, Dorr U. Cancer pain man-agement according to WHO analgesic guidelines. JPain Symptom Manage 1990;5:2732.</p><p>4. Zech DFJ, Grond S, Lynch J, Hertel D, LehmanKA. Validation of World Health Organisation guide-lines for cancer pain relief. A 10-year prospectivestudy. Pain 1995;63:6576.</p><p>5. Hanks G, Portenoy RK, MacDonald N, Forbes K.Difficult pain problems. In: Doyle D, Hanks GWC,MacDonald N, eds. Oxford textbook of palliativemedicine, 2nd ed. Oxford: Oxford University Press1998: 454477.</p><p>6. Mercadante S, Maddaloni S, Roccella S, Salvag-gio L. Predictive factors in advanced cancer paintreated only by analgesics. Pain 1992;50:151155.</p><p>7. Po...</p></li></ul>


View more >