Antibiotic Use and Effects in the Terminally Ill : – A Retrospective Review

(1)

Antibiotic Use and Effects in the Terminally Ill

– A Retrospective Review

Version 2

Author: Ida Wikblom

Örebro University, Sweden Bachelor in Medicine

Supervisor: Åsa Andersson, RTP, PhD

Örebro University Hospital, Sweden Department of Geriatric Medicine Örebro University

School of Medicine Degree Project – 30 ECTS

November 2017 Sweden

(2)

Abstract

Introduction: The use of antibiotics on palliative patients is greatly varying around the globe. Studies on the beneficial effects of antibiotics are scarce and conflicting, as is the connection between antibiotics and survival.

Aim: To survey the use and effects of antibiotics in palliative patients. We aimed to

investigate the effect of antibiotics when comparing different sites of infection and to study change in CRP during antibiotic administration and survival after end of treatment.

Method: A retrospective review including all patients deceased during 2016 while admitted to the palliative unit in Örebro. Specifics concerning antibiotics were collected if the patient had ongoing treatment during the last two weeks in life. Symptom relief, elevated general

condition or functional level was considered positive effects of antibiotics.

Result: Of the total 191 patients reviewed, 73 met the inclusion criteria for the antibiotic group. Overall positive effect was seen in 42% of treatment episodes. Highest positive effect was received when treating skin infections (7/7). Second and third largest effect was seen in the “sepsis” (11/20) and “respiratory tract” (9/20) groups. An additional four days of median survival in favour of CRP decline (p = 0.04) was identified when compared rising trends.

Conclusion: A trend towards longer median survival in patients responding with CRP decline compared to CRP rise during antibiotic therapy was observed. The positive effect was

however spread between different infection sites. A time-limited trial of antibiotic therapy might be preferable since it’s hard to foretell which patients will receive beneficial effects.

(3)

Introduction

Palliative care

When patients are terminally ill and no curative treatment is longer available the tending converts to palliative care [1]. The practicing of palliative care can be separated into three main subgroups consisting of specialized palliative care, general palliative care and hospice care. Specialized palliative care is further divided into either hospital based or advanced home care. For admittance to a specialized unit the patients usually require a more complex

symptomatology, often associated with cancer disease [2,3].

Palliative care is truly accomplished when focus remains on the patient and their families’ total well-being. The well-being as composed of physiological, psychological and spiritual quality of life as opposed to survival being the central outcome [1]. As a result of this core motivation, decision-making concerning the treatment of comorbidities diverge from the general practice. Evaluation of the potential profit, adverse effects and possible life sustaining quality of the discussed intervention must be regarded and communication between health care provider and patient is essential [4,5].

Antibiotics and Infections

Infections are common in the dying in-house patient and can often be linked to cause of death [6]. The use of antibiotics on palliative patients is however greatly varying around the globe and in different palliative settings. According to a systematic review from 2013 numbers are ranging from 8-84% [7]. Two recent studies restricted to in-house cancer patients during their last weeks of life shows more constringent numbers around 50% [8,9].

Written guidelines on when to treat suspected infections are non-existent and consensus on the subject is lacking [4,10]. The topic of antibiotic treatment is therefore regularly discussed among the palliative medical staff and the opinions often diverge. Studies on the beneficial effects of antibacterial agents on palliative patients are scarce and conflicting [8,11–16] as is the connection between antibiotics and survival [11,14,17,18]. However, since survival is only secondary to symptomatic relief when performing palliative care and due to the ongoing debate, the subject is important to further illuminate. We conducted this study to survey the use of antibiotics and determine its effect during the last two weeks of life in patients deceased while receiving specialized palliative care in Örebro during 2016.

(5)

Aim

The primary aim was to survey the characteristics of antibiotic use in palliative patients during their last weeks of life. The secondary aims were A) to investigate if there was a difference in measured effect of antibiotics when comparing treatment episodes divided into suspected site of infection and B) to study the change in CRP during antibiotic administration and survival after end of treatment.

Material and methods

Study design and cohort

To elucidate the effect of antibiotics near end of life we conducted this retrospective study enrolling all patient deceased while admitted to the specialized palliative section at the department of Geriatrics at Örebro University Hospital during 2016. This included all patient passing away either in the hospital ward or at home with the aid of ASIH (advanced palliative home care). All data was collected from electronic journals using a standardized template and thereafter assembled in Microsoft Excel (version 15.40). Baseline data was obtained from all patients. Specifics concerning antibiotic treatment and infection was collected from those receiving antibiotics during the last two weeks of life. The effect of antibiotics was classified as positive or negative. Positive effect considered coherent with antibiotic treatment was either symptom relief, elevated general condition, raised functional level or other. The negative category was applied when no positive effect was identified or if the patient

experienced side effects. Lab values of C-Reactive Protein (CRP) was recorded. The values in closest proximity to initiation and end of treatment was collected. If existing, a top notation and lowest CRP during treatment was also obtained. This enabling calculation on either declining or rising trends.

Statistics

T-test was used to analyse the difference in age and hospitalizations. Chi-squared tests for analysing differences in proportions for dichotomized variables. The difference in number of days to death was analysed by using Mann-Whitney’s test. The level of statistical significance was set at p < 0.05. The survival based on CRP trend has been illustrated with Kaplan-Meier graphs. All data was analysed using IBM SPSS Statistics (version 22).

Ethics

(6)

geriatric clinic at Örebro University Hospital. According to Swedish law no official ethical approval was hence required [19]. The study was however conducted under ethical and safe conditions in accordance with the Declaration of Helsinki. All included patients was given a code number and all data was presented at group level [20].

Results

Patient characteristics

A total of 191 unique patients was enrolled. The median length of stay at the palliative unit was 9 days (IQR 3-22). The four most common diagnosis was lung (n = 31)(16%), pancreas (n = 19)(10%), colon (n = 17)(9%) and prostate cancer (n = 16)(8%). Twenty-two (12%) patients were categorized as “no cancer”. These patients suffered from a variety of different diagnoses such as major bleeding events, sepsis, heart and kidney failure. The characteristics of all patients are presented in Table 1.

Table 1. Characteristics of patients deceased while admitted to a specialized palliative care unit during 2016, n =

191.

All patients Cancer patients

Received antibiotics

Total 191 169 73

Female, n (%) 89 (47) 80 (47) 33 (45)

Age, Mean years (SD) 73 (11) 72 (10) 73 (11)

Type of palliative care*, n (%)

ASIH** 56 (29) 55 (33) 24 (33)

General 47 (25) 42 (25) 18 (25)

None 88 (46) 72 (43) 31 (43)

Hospitalizations during last year of life, Mean n (SD) Range 4 (2) 0-14 4 (2) 0-14 4 (2) 2-11 Place of death, n (%)

Palliative hospital ward 173 (91) 152 (90) 64 (88)

At home with ASIH** 18 (9) 17 (10) 9 (12)

(7)

Figure 1. Distribution of antibiotics between different patient groups with and without cancer, n = 191.

Seventy-three patients were treated with antibiotics during their last two weeks in life (patients with cancer n = 63, without cancer n = 10). Patients suffering from

esophagus/gastric and pancreatic cancer had the highest and lowest frequency of antibiotic use, 53% and 16% respectively (see Figure 1 for all subgroups presented). Contrasting “cancer” and “no cancer” patients the frequencies were 37% and 45%.

Comparing the antibiotic and non-antibiotic group, no significance regarding age (mean years 73 and 72 respectively, p = 0.51) or number of hospitalizations during the last year in life was observed (mean hospitalizations four in both groups, p = 0.72). The non-cancer patients were however significantly older (mean years 81 and 71 respectively; p < 0.001) than the cancer patients but no difference was noted regarding hospitalization (mean hospitalizations four in both groups, p = 0.74).

Aspects of treatment and effects

Five patients received antibiotics two times during this period leaving us with 78 unique suspected infections treated (for characteristics see Table 2). Five of these 78 treatment episodes appeared without recording of time for initiation and/or measurable effect representing parts of the missing data in this study. Among these five, prophylaxis and longstanding treatment was the main indications.

0 2 4 6 8 10 12 14 16 18 20 Nu m b er of pati en ts No Antibiotics Antibiotics

(8)

Table 2. Characteristics of treatment episodes ongoing the last two weeks in life, n = 78.

All No cancer

Suspected infections treated with antibiotics n 78 10

Place of initiation n

Treatment initiated by Palliative unit 46 0

Treatment initiated by Other unit 19 8

Treatment overtaken by Palliative unit 13 2

Infection specialist consultation n 24 4

Reason for terminating antibiotic treatment n

Treatment completed 17 2

Will of patient 5 0

No effect 14 3

Reduced general condition 14 2

Death 8 0

Other 7 2

Not documented 13 1

Days of treatment, Md (IQR) Range (n = 76)* 6 (3-11) 0-78 12 (2-22) 0-32

*Missing data on two treatment episodes.

Of the 78 episodes treated, a positive effect was achieved in 33 cases. A decline in CRP was subsequently observed in 22 of these cases. CRP decline was in addition seen in 18 patients categorized as receiving a negative effect. The received effect was not significantly associated to CRP trend (p = 0.94). No side effects were reported.

A total of 106 bacterial cultures was taken, these distributed among 51 treatment episodes. Thirty-eight episodes had one or more positive culture (in total 51/106 positive cultures). Blood and urine samples was most commonly obtained, constituting for 38 and 39 cultures respectively.

Overall, having a positive culture was not significantly associated to receiving either a

positive or negative effect of antibiotic treatment (p = 0.78). Neither when compared to rising or falling CRP trends (p = 0.15). The most common pathogen was Escherichia coli detected in 6/15 positive blood samples and 8/22 positive urine samples. ESBL (extended spectrum beta lactamase) was found in four cultures collected from two unique patients. The highest frequency of positive tests was observed in skin cultures (6/7) with predominantly different coliform bacteria.

(9)

Suspected site of infection and effect

When comparing the effect between groups of “suspected site of infection” we found a 100% positive effect when treating suspected skin infections, the measured effect was symptom relief in all treatment episodes. The second highest positive effect was measured in the “sepsis” group residing on 55% followed by “respiratory tract”, 45% (Table 3). Elevated general condition and/or functional level was the central effects compared to symptomatic relief when treating sepsis and respiratory tract infections respectively. CRP decline or rise was widely spread between all groups (Figure 2), as was the connection to positive cultures.

Table 3. Suspected site of infection and effect of antibiotic treatment, n = 78.

Suspected site of infection Total n (%) Effect measured* n

Positive effect n (% of effect measured)

Positive effect associated decline in CRP n (missing data n) All 78 (100) 73 33 (45) 22 (4) Sepsis 20 (26) 20 11 (55) 9 (1) Respiratory tract 21 (27) 20 9 (45) 8 Urinary tract 14 (18) 13 3 (23) 0 (2) GI tract/abdomen 9 (12) 8 2 (25) 1 Skin 7 (9) 7 7 (100) 3 (1) Other 5 (6) 4 1 (25) 1 Not documented 2 (3) 1 0 (0) 0 *Missing data, n = 5.

Figure 2. Suspected site of infection and CRP change during antibiotics administration. Treatment episodes, n =

78. 0 5 10 15 20 25 Respiratory Sepsis Urinary Abdominal Skin Other Not documented

CRP decline < 50 CRP decline 50-100 CRP decline 101-150 CRP decline > 150 CRP rise Data missing

(10)

CRP change and survival

To investigate if there was a difference between change in CRP during treatment and time from termination of treatment until death we divided the values (data available n = 54) in either CRP rise or CRP decline. As seen in Figure 3 the median survival was an additional four days in favour of CRP decline (CRP decline: Median 6 days, 95% CI 5-7; CRP rise: Median 2 days, 95% CI 0-5) (p = 0.04).

Discussion

Terminally ill patients are at high risk of acquiring infections and the decisions on treatment initiation, withholding or withdrawal must be gently balanced to fit each patients’ treatment goals. Our study illustrates the complexity of treating palliative patients due to the diversity of disease and different treatment goals. This resulting in largely deviating data, difficulty in creating useful tools for decision making and in finding consensus on the subject.

Number of days after end of treatment

C u m u la tiv e S u rvi val

Figure 3. Comparing change in CRP during antibiotic administration and days of survival from

(11)

Our study is covering a full year of treated patients and all journals has been reviewed with a standardized template by the author alone. We chose to include all patients regardless of primary diagnosis to expand our data, consequently limiting the comparability with studies on cancer patients alone. We also included treatment both ongoing and initiated during the last two weeks life which resulted in comparing treatment episodes with greatly varying lengths. This further compromising the generalization as well as lowering the internal validity of our study. Nonetheless, the amount of obtained data was still considerably small.

Overall antibiotic use in our cohort was 38% (cancer 37%, no cancer 45%), slightly below the frequency presented by Helde-Franklin et al (49%), a recent retrospective study from Sweden restricted to 160 in house or home based palliative cancer patients during their last two weeks in life [8]. The use is however greatly varying worldwide and in different palliative settings due to different attitudes and patient characteristics [5,7,21]. Yao et al presented that the most influential factor on palliative patients’ wishes considering antibacterial agents was medical staff opinions. “Antibiotic use is helpful to all terminal patients with infection”, was the dominating misconception [5].

Escherichia coli was the most common pathogen isolated in cultures in this study. This in line with previous studies [13,14,21]. The presence of ESBL was comparable with the frequency presented by Helde-Frankling et al [8] and correlates well with the relatively low occurrence of multi resistant bacteria in Sweden [22].

A positive effect was observed in 45% of treatment episodes. Similar response was reported by two comparable retrospective studies concerning treatment during the last admission or last two weeks in life (37% and 40%) [8,14]. Interestingly, Oh et al, studying palliative cancer patients, found that merely 15% of patients achieved symptomatic improvement but 48% of all fevers was resolved [12]. Since paraneoplastic fever is common in advanced cancer and can be difficult to differentiate from infection [23] this is an important factor to consider before initiating treatment since it might decrease the amount of unnecessary use.

The generally high occurrence of negative effect and the weak association between effect and having positive culture may be explained by the prominent severity of illness present in our cohort, consequently resulting in an inability to be assisted by antibiotic treatment.

(12)

prolongation of life due to antibiotics are effects that might be camouflaged in the negative effect group. Also, important to consider when analyzing the measured effect is the

retrospective nature of this study. Collection of clinical features concerning infections from written journal texts might underestimate the effect of treatment. However, measuring the effect by identifying specific clinical improvement areas can instead result in an

overestimation of the beneficial influence. For example, a raised functional level can just as well be related a better pain management and not an effect of infection control.

In our study, the greatest amount of positive effect was measured in the small skin foci group, followed by sepsis and respiratory tract infections. In contrast, less than half of the skin infections (44%) but the majority of urinary- (83%) followed by respiratory (40%) tract infections were symptomatically controlled in a prospective study on 255 late stage cancer patients receiving outpatient hospice care [13]. These observations were also supported by other studies with varying study designs [14–16]. Notably, bacteraemia received a generally poor response in all these studies and did only represent small groups. For inclusion into this category the patient did however require a positive blood culture. No suspected sepsis groups were presented. The results can hence not be directly compared to ours, where suspected sepsis was solely based on the first clinical evaluation documented.

Helde-Franklin et al presented a significant difference when comparing symptom relief between sepsis and urinary tract infection, favouring the treatment of sepsis. Symptom relief was seen in 50% and 17% respectively [8]. This comparable to our results were a positive effect was measured in 55% of suspected sepsis and 23% of suspected urinary tract foci. These findings are however diverging from the results presented above and since

sepsis/bacteraemia was otherwise defined, miscategorised positive effect of actual urinary tract infections might be hidden in the sepsis group.

Receiving a measurable positive effect was not related to a decline in CRP. However, to investigate whether a change in CRP could influence survival we separated the CRP trends into different groups. Our findings indicate that a decline in CRP may be favorable for

survival after end of treatment compared rising trends. Nevertheless, it is important to keep in mind that the value of CRP can be affected by many factors, not only infection. For example, treatment episodes with rising trends might as well have infections present and an effect of

(13)

CRP as a tool for identifying inflammation is generally accepted and the use for monitoring infection will most likely continue indefinite. Penel et al presented that CRP failed to

discriminate infection from paraneoplastic fever but high levels of CRP was closely correlated to death when infection was present [24]. This, in line with our findings, also demonstrating CRP as a potential instrument when discussing antibiotic treatment.

To our knowledge, there are no additional comparable data on antibiotic treated infections, CRP and survival concerning palliative patients. However, the connection between use of antibiotics or presence of infection/infective symptoms and survival has been addressed in a couple of publications [11,14,17,18]. Findings made by Thai et al studying late stage cancer patients suggested that a good initial response to treatment was associated with an increase in median survival (72 days) when compared a poor initial response (19 days). Response was measured in either symptom resolution, changes in blood works or imaging [17]. These results can to a limited extent be mirrored in our findings but the resemblances in study design are minimal.

The continued evolvement of palliative medicine is in need of further research. The studying of antibiotics and its clinical effects in palliative patients might however benefit from new angles of approach. The research on antibiotics and its effect on specific sites of infections has been on timeout during the last 10 years. This might be explained by the previous studies inability to generate consensus on the subject. Accordingly, we also found that the beneficial effect was spread between these groups. We did however observe a trend towards longer survival after termination of treatment in patients responding with CRP decline compared CRP rise during antibiotic therapy. Nevertheless, this association needs to be confirmed in larger prospective studies.

Conclusion

Our study indicates that a time-limited trial of antimicrobial therapy monitoring the effect on symptoms and CRP might be preferable since it is hard to foretell what patients will receive beneficial effects. The choice of initiation, withholding and withdrawal must nonetheless be thoroughly discussed with the affected individuals and the potential outcomes

(14)

References

1. WHO | WHO Definition of Palliative Care [Internet]. WHO. [citerad 28 augusti 2017]. Tillgänglig vid: http://www.who.int/cancer/palliative/definition/en/

2. Vad är palliativ vård? - Nationella Rådet för Palliativ Vård | Nationella Rådet för Palliativ Vård [Internet]. [citerad 30 augusti 2017]. Tillgänglig vid:

https://www.nrpv.se/vad-ar-palliativ-vard/

3. Dhingra L, Barrett M, Knotkova H, Chen J, Riggs A, Lee B, m.fl. Symptom Distress Among Diverse Patients Referred for Community-based Palliative Care:

Sociodemographic and Medical Correlates. J Pain Symptom Manage. 24 augusti 2017; 4. van Nordennen RTCM, Lavrijsen JCM, Vissers KCP, Koopmans RTCM. Decision making about change of medication for comorbid disease at the end of life: an integrative review. Drugs Aging. juli 2014;31(7):501–12.

5. Yao C-A, Hsieh M-Y, Chiu T-Y, Hu W-Y, Hung S-H, Chen C-Y, m.fl. Wishes of Patients With Terminal Cancer and Influencing Factors Toward the Use of Antibiotics in Taiwan. Am J Hosp Palliat Care. augusti 2015;32(5):537–43.

6. Abdel-Karim IA, Sammel RB, Prange MA. Causes of death at autopsy in an inpatient hospice program. J Palliat Med. augusti 2007;10(4):894–8.

7. Rosenberg JH, Albrecht JS, Fromme EK, Noble BN, McGregor JC, Comer AC, m.fl. Antimicrobial Use for Symptom Management in Patients Receiving Hospice and

Palliative Care: A Systematic Review. J Palliat Med. 01 december 2013;16(12):1568–74. 8. Helde-Frankling M, Bergqvist J, Bergman P, Björkhem-Bergman L. Antibiotic Treatment in End-of-Life Cancer Patients-A Retrospective Observational Study at a Palliative Care Center in Sweden. Cancers. 06 september 2016;8(9).

9. Abduh Al-Shaqi M, Alami AH, Zahrani ASA-, Al-Marshad B, Muammar AB-, M Z A-S. The pattern of antimicrobial use for palliative care in-patients during the last week of life. Am J Hosp Palliat Care. februari 2012;29(1):60–3.

10. Juthani-Mehta M, Malani PN, Mitchell SL. Antimicrobials at the End of Life: An Opportunity to Improve Palliative Care and Infection Management. JAMA. 17 november 2015;314(19):2017–8.

11. Chen L-K, Chou Y-C, Hsu P-S, Tsai S-T, Hwang S-J, Wu B-Y, m.fl. Antibiotic prescription for fever episodes in hospice patients. Support Care Cancer Off J Multinatl Assoc Support Care Cancer. oktober 2002;10(7):538–41.

12. Oh DY, Kim JH, Kim DW, Im SA, Kim TY, Heo DS, m.fl. Antibiotic use during the last days of life in cancer patients. Eur J Cancer Care (Engl). mars 2006;15(1):74– 9.

13. White PH, Kuhlenschmidt HL, Vancura BG, Navari RM. Antimicrobial use in patients with advanced cancer receiving hospice care. J Pain Symptom Manage. maj

2003;25(5):438–43.

14. Vitetta L, Kenner D, Sali A. Bacterial Infections in Terminally Ill Hospice Patients. J Pain Symptom Manage. 01 november 2000;20(5):326–34.

15. Reinbolt RE, Shenk AM, White PH, Navari RM. Symptomatic treatment of infections in patients with advanced cancer receiving hospice care. J Pain Symptom Manage. augusti 2005;30(2):175–82.

16. Clayton J, Fardell B, Hutton-Potts J, Webb D, Chye R. Parenteral antibiotics in a palliative care unit: prospective analysis of current practice. Palliat Med. januari

2003;17(1):44–8.

17. Thai V, Lau F, Wolch G, Yang J, Quan H, Fassbender K. Impact of infections on the survival of hospitalized advanced cancer patients. J Pain Symptom Manage. mars

(15)

18. Chih A-H, Lee L-T, Cheng S-Y, Yao C-A, Hu W-Y, Chen C-Y, m.fl. Is it appropriate to withdraw antibiotics in terminal patients with cancer with infection? J Palliat Med. november 2013;16(11):1417–22.

19. Regeringskansliet R och. Etikprövning av forskning [Internet]. Regeringskansliet. 2003 [citerad 07 oktober 2017]. Tillgänglig vid:

http://www.regeringen.se/rattsdokument/proposition/2003/01/prop.-20020350/ 20. World Medical Association. World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects. JAMA. 27 november 2013;310(20):2191–4.

21. Albrecht JS, McGregor JC, Fromme EK, Bearden DT, Furuno JP. A nationwide analysis of antibiotic use in hospice care in the final week of life. J Pain Symptom Manage. oktober 2013;46(4):483–90.

22. Antimicrobial resistance surveillance in Europe 2015 [Internet]. European Centre for Disease Prevention and Control. 2017 [citerad 31 oktober 2017]. Tillgänglig vid: http://ecdc.europa.eu/en/publications-data/antimicrobial-resistance-surveillance-europe-2015 23. Zell JA, Chang JC. Neoplastic fever: a neglected paraneoplastic syndrome. Support Care Cancer Off J Multinatl Assoc Support Care Cancer. november

2005;13(11):870–7.

24. Penel N, Fournier C, Clisant S, N’Guyen M. Causes of fever and value of C-reactive protein and procalcitonin in differentiating infections from paraneoplastic fever. Support Care Cancer Off J Multinatl Assoc Support Care Cancer. augusti 2004;12(8):593–8.

(16)

Ethical Consideration

This study was conducted to review the use and effects of antibiotic treatment on the

terminally ill. This is an important subject to highlight since antibiotics in some respects can be considered an invasive treatment when terminally ill. The psychological part of dying and acceptance of death approaching can be obstructed by continued interventions and hope for turning of the tides.

The central ethical aspect when performing studies on already deceased patients is the inability to obtain informed consent. However, the presentation of data in our study is designed for the patients’ fullest discretion. All included patients were de-identificated by receiving a code number and all data was presented at group level, this in line with the Declaration of Helsinki.

Another aspect, concerning the collection of data retrospectively, is the intrusion into journals and related sensitive information. When reviewing journals, it is impossible not to acquire additional data not associated with the present study. Furthermore, retrospective studies might not be considered invasive, except this personal intrusion, but still do not benefit the study subjects. Also an ethical dilemma.

Nevertheless, in favor of this retrospective nature, the patients included are not subjected to any risks or burdens. There is no intervention affecting their bodies, no unwanted side effects to be caused and no other interference with their lives, neither physically nor psychologically. The potential beneficial findings applicable on future patients must therefore be considered superior the inability of obtaining informed consent.

(17)

Cover Letter

December 11th, 2017. Corresponding author: Ida Wikblom, Bachelor of medicine, Örebro University.

Dear Editor

Please let us present our manuscript “Antibiotic Use and Effects in the Terminally ill – A Retrospective Review”, to be considered for publication as an original article in the Journal of palliative medicine.

In our study, we surveyed the use and effects of antibiotics among palliative patients during their last weeks in life. Our findings suggest that CRP might act as support in the decision-making process concerning withdrawal or withholding of antibiotic treatment since a decline in CRP during treatment showed a trend towards longer survival after end of treatment. When studying the clinical effect compared between different infections foci the measured effect was spread among the groups. The research on antibiotics and its effect on specific sites of infections has been on timeout during the last 10 years. This might be explained by the previous studies inability to generate consensus on the subject, also adressed in our study.

As you know, the treatment of palliative patients is a delicate subject with a lot of ethical considerations. Opinions concerning treatment and interventions are often deviating among health care personnel, patients and relatives. Our study is important since it highlights the complexity of treating the terminally ill and might act as groundwork for future research in the area of palliative medicine.

This manuscript has not been published elsewhere.

Best Regards, Ida Wikblom

(18)

Antibiotika till den döende patienten – Lindring eller förlängt lidande?

Vården kring den döende patienten har en annan inriktning än övrig vård. Fokus ligger på ett totalt välmående där kroppsliga och själsliga komponenter är jämställda. Besluten kring

läkemedelsbehandling blir därav mer komplexa. Frågan om antibiotika är gynnande i livets slutskede har ännu ej med säkerhet besvarats, heller ej om det förlänger livet.

Vår studies målsättning var att belysa just detta område. Vi ville undersöka om en skillnad i effekt av antibiotika kunde identifieras beroende på vilken typ av infektion som behandlades. Vi önskade även att analysera ifall en förändring av snabbsänkan* under aktuell behandling kunde kopplas till längre överlevnad.

Sjuttiotre av totalt 191 granskade patienter erhöll antibiotika under de två sista

veckorna i livet. Fyrtiofem procent hade en mätbar positiv effekt av behandlingen.

Patienterna hade olika grundsjukdomar men majoriteten utgjordes av cancer.

När vi undersökte effekten av antibiotika mellan olika sorters infektioner kunde vi inte hitta någon betydelsefull skillnad dem emellan. Däremot fann vi en längre

överlevnad efter utsatt behandling bland de patienter som hade sjunkande värden av snabbsänka jämfört med stigande värden under pågående behandling.

Vår studie belyser det faktum att patienter i livets slutskede är väldigt unika vilket följaktligen skapar svårigheter med

generalisering av vården. Snabbsänkan kan dock vara till hjälp vid beslut om

utsättning av antibiotika. Det viktigaste redskapet är emellertid kommunikation mellan patienten, anhöriga och

vårdpersonal. Detta för att tidigt skapa gemensamma behandlingsmål som kan vägleda i svåra beslut den sista tiden i livet. __________________________________

* Snabbsänkan är ett labbvärde som kan spegla en pågående infektion i kroppen.

Antibiotic Use and Effects in the Terminally Ill : – A Retrospective Review