Association between hydration volume and symptoms in terminally ill cancer patients with abdominal malignancies

T. Morita1,*, I. Hyodo2, T. Yoshimi3, M. Ikenaga4, Y. Tamura5, A. Yoshizawa6, A. Shimada7, T. Akechi8, M. Miyashita9, I. Adachi10 Japan Palliative Oncology Study Group{dagger}

1 Department of Palliative and Supportive Care, Palliative Care Team and Seirei Hospice, Seirei Mikatabara Hospital, Hamamatsu, Shizuoka; 2 Gastroenterology, National Shikoku Cancer Center Hospital; 3 Internal Medicine, Dozono Medical House; 4 Hospice, Yodogawa Christian Hospital; 5 Department of Surgery, Kasumigaura National Hospital; 6 Internal Medicine, Kanamecho Hospital; 7 Department of Palliative Medicine, Tohoku University Hospital; 8 Psycho-Oncology Division, National Cancer Center Research Institute East, Psychiatry Division, National Cancer Center Hospital East; 9 Department of Adult Nursing/Terminal and Long-term care Nursing, The University of Tokyo; 10 Department of Palliative Medicine, Shizuoka Cancer Center, Japan

* Correspondence to: Dr T. Morita, Department of Palliative and Supportive Care, Palliative Care Team and Seirei Hospice, Seirei Mikatabara Hospital, 3453 Mikatabara-cho, Hamamatsu, Shizuoka 433-8558, Japan. Tel: +81-053-436-1251; Fax: +81-053-438-2971; Email: seireihc{at}jt6.so-net.ne.jp


    Abstract
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Background: To explore the association between hydration volume and symptoms during the last 3 weeks of life in terminally ill cancer patients.

Patients and methods: This was a multicenter, prospective, observational study of 226 consecutive terminally ill patients with abdominal malignancies. Primary responsible physicians and nurses evaluated the severity of membranous dehydration (dehydration score calculated from three physical findings), peripheral edema (edema score calculated from seven physical findings), ascites and pleural effusion (rated as physically undetectable to symptomatic), bronchial secretion, hyperactive delirium (Memorial Delirium Assessment Scale), communication capacity (Communication Capacity Scale), agitation (Agitation Distress Scale), myoclonus and bedsores.

Results: Patients were classified into two groups: the hydration group (n=59) who received 1 l or more of artificial hydration per day, 1 and 3 weeks before death, and the non-hydration group (n=167). The percentage of patients with deterioration in dehydration score in the final 3 weeks was significantly higher in the non-hydration group than the hydration group (35% versus 14%; P=0.002), while the percentages of patients whose symptom scores for edema, ascites and pleural effusion increased were significantly higher in the hydration group than the non-hydration group (44% versus 29%, P=0.039; 29% versus 8.4%, P <0.001; 15% versus 5.4%, P=0.016; respectively). After controlling for multiple covariates and treatment settings, the association between hydration group and dehydration/ascites score was statistically significant. Subgroup analysis of patients with peritoneal metastases identified statistically significant interaction between hydration group and dehydration/pleural effusion score. There were no significant differences in the degree of bronchial secretion, hyperactive delirium, communication capacity, agitation, myoclonus or bedsores.

Conclusions: Artificial hydration therapy could alleviate membranous dehydration signs, but could worsen peripheral edema, ascites and pleural effusions. It is suggested that the potential benefits of artificial hydration therapy should be balanced with the risk of worsening fluid retention symptoms. Further clinical studies are strongly needed to identify the effects of artificial hydration therapy on overall patient well-being, and an individualized treatment and close monitoring of dehydration and fluid retention symptoms is strongly recommended.

Key words: dehydration, neoplasm, palliative care, rehydration, water depletion


    Introduction
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
The dehydration–rehydration problem has been one of the most important issues in palliative or end-of-life care literature over the two last decades [1Go]. Current discrepancies in the practice of artificial hydration therapy for terminally ill cancer patients have the potential to cause serious clinical problems: patients could suffer from unnecessary dehydration-related symptoms or experience iatrogenic over-hydration symptoms [2Go–5Go]. These discrepancies are largely due to the lack of evidence about the effects of artificial hydration therapy on patient well-being [3Go].

Traditionally, artificial hydration therapy has been thought not to benefit the terminally ill [6Go–11Go]; however, some recent studies have demonstrated that appropriate hydration can contribute to patient comfort [12Go–16Go]. The majority of studies on this topic are limited by methodological issues [17Go], and do not provide enough of a basis for the evidence-based practice of artificial hydration therapy in terminally ill patients. The aim of the present study was to explore systematically the associations between hydration volume and dehydration and fluid retention symptoms in the last 3 weeks of life in terminally ill patients with abdominal malignancies.


    Patients and methods
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Patients
The study subjects were consecutive terminally ill cancer patients treated in 14 oncology units, 19 palliative care units and four home-based palliative care programs in Japan. The participating institutions recruited potential participants following the same inclusion criteria: age >20 years; life expectancy estimated by a physician to be ≤3 months; and incurable malignancy of abdominal origin (excluding hepatic malignancies). Exclusion criteria were: liver cirrhosis of any etiology, renal failure, nephrotic syndrome, protein-losing enteropathy, intra-abdominal shunt for ascites, hypercalcemia, adrenalopathy, thyroid diseases, and other complications of the circulatory, respiratory, hepatic, or renal system unrelated to underlying malignancies; surgical, radiological or oncological treatments with the primary intent of tumor reduction in the 3 weeks prior to study inclusion; existing communication difficulty such as aphasia or aphonia; and the use of artificial enteral nutrition. Patients were enrolled from August 2002 to February 2003, and followed until March 2003.

Study design
This was a multicenter, prospective, observational study. From the time of study inclusion, primary responsible physicians prospectively recorded patients' dehydration and fluid retention symptoms on a structured data-collecting sheet every week as a part of daily practice. In addition, symptoms observed very close to death were assessed after patients died, because it was impossible to predict when the patients would die, and because assessments on a daily basis would present a high burden for patients and physicians. Thus, within 72 h after patient death, patients' dehydration and fluid retention symptoms 24 h before death, communication capacity during the 3 days before death and degree of agitation in the week before death were recorded. To minimize the recall bias, the evaluations were based on the full agreement of the primary physicians and primary nurses.

The patients received the usual treatments from their institutions. The indications for hydration therapy, administration methods and modification of treatment regimen over time were dependent on each physician's clinical decisions. We did not standardize hydration treatments, because patients' and families' wishes and each physician's philosophy strongly influenced actual hydration practice [3Go], and adopting a single hydration protocol was regarded as inappropriate outside experimental study designs. Instead, we described the details of hydration treatments actually performed for interpretation of the results.

This study was approved by the Institutional Review Board of each hospital, and conducted in accordance with the Declaration of Helsinki.

End points and measurements
The primary end points of this study were dehydration and fluid retention symptoms in the last 3 weeks of life. Although patient-reported symptoms and satisfaction are important outcomes in palliative care [18Go], we chose symptoms that could be objectively evaluated as the end points for this study. The rationale for this decision was that adopting self-report measures could result in higher rates of patient exclusion and unacceptable selection bias, because patient reports are often impossible in the very late stages of cancer due to cognitive impairment, and because symptom evaluations based on patient self-reports are not routinely used in many participating institutions [19Go, 20Go].

Physical symptoms
Physicians were requested to perform physical examinations in the morning at least 1 h after patients had eaten. The degree of dehydration was assessed on the basis of three physical findings: moisture on the mucous membranes of the mouth (0, moist; 1, somewhat dry; 2, dry), axillary moisture (0, moist; 1, dry) and sunkenness of eyes (0, normal; 1, slightly sunken; 2, sunken). These signs were selected due to their significant correlations with biological dehydration, as previously confirmed in elderly patients [21Go–23Go]. Empirical studies have found that the sensitivity/specificity of each sign in identifying dehydration is 85%/58%, 50%/82% and 62%/82%, respectively [21Go–23Go]. Ad hoc dehydration score (range 0–5) was calculated as the total of these three scores. A higher score thus indicated a higher level of dehydration.

The severity of peripheral edema was determined through the examination of seven regions: the hands, forearms, upper arms, feet, lower legs, thighs and trunk. Peripheral edema severity was scored based on the degree of increased skin thickness in the middle of each region (0, none; 1, mild, thickness of <5 mm; 2, moderate, 5–10 mm; 3, severe, >10 mm). If peripheral edema was asymmetric, the more severe side was rated unless the asymmetry was caused by a unilateral vascular obstruction; in these cases, the non-obstructed side was rated. The peripheral edema score (range 0–21) was calculated as the total of the severity scores for the seven regions. A higher score indicated more severe edema.

Pleural effusion and ascites were each rated on a scale of 0 to 2 (0, physically non-detectable; 1, physically detectable but asymptomatic; 2, symptomatic or tense ascites). Myoclonus and bedsores were considered present when they were observed at any time in the final 3 weeks of life.

Bronchial secretion was defined as sounds audible at the bedside produced by movement of secretions in the hypopharynx or the bronchial tree in association with respiration [8Go]. The severity of bronchial secretion was evaluated using a previously proposed scale: ‘inaudible’ (score 0), ‘audible only very close to the patient’ (score 1), ‘clearly audible at the end of the bed in a quiet room’ (score 2) and ‘clearly audible at about 6 m or at the door of the room’ (score 3) [24Go]. Bronchial secretions were considered present when patients had a severity score of 1 or more, received any anti-muscarinic medications to reduce bronchial secretion or received oral/bronchial suctioning at least once during the final 3 weeks of life. Severe bronchial secretion was defined as severity score of 2 or 3 at any time during the final 3 weeks.

Psychiatric symptoms
We used selected items of the Communication Capacity Scale, the Agitation Distress Scale and the Memorial Delirium Assessment Scale to evaluate psychiatric symptoms [25Go, 26Go]. The Communication Capacity Scale is a validated five-item observer-rating scale used to quantify communication capacity in terminally ill patients [25Go]. The Agitation Distress Scale is a six-item observer-rating scale used to quantify the levels of agitation in delirious terminal patients [25Go]. Although using all items of a scale is psychometrically ideal, we used select items in order to reduce physician burden and increase patient enrollment [27Go].

The patients' communication capacity was assessed using the highest scores measured in the last 3 days of life on three items: the ‘reduced level of consciousness’ item of the Memorial Delirium Assessment Scale, and the ‘answers to closed-ended questions’ and ‘voluntary communication’ items from the Communication Capacity Scale. Communication score (range 0–9) was calculated as the total of these three items, such that a higher score indicated a greater capacity for communication (Cronbach's {alpha} coefficient = 0.94). The correlations between total score on this abbreviated scale and the total score on the Communication Capacity Scale was high in the original validation data (Spearman's {rho}=0.94; P <0.001) [25Go].

The degree of agitation was defined as the most severe symptoms experienced during the last week of life, and quantified using four items from the Agitation Distress Scale: the frequency of motor anxiety, extent of motor anxiety, contents of motor anxiety and psychological instability. The agitation score (range 0–12) was calculated as the total of these four items; a higher score indicated higher levels of agitation (Cronbach's {alpha} coefficient = 0.87). The correlations between total score on this abbreviated scale and the total score on the Agitation Distress Scale was high in the original validation data (Spearman's {rho}=0.95; P <0.001) [25Go].

Hyperactive delirium was assessed using the ‘psychomotor activity’ item of the Memorial Delirium Assessment Scale, which grades increased psychomotor activity on a scale of 0 (normal) to 3 (severe) [26Go]. Hyperactive delirium was defined as a score of 2 or 3 on this scale.

Covariates of main outcomes
We recorded the presence or absence of the following potential covariates: stomatitis, oxygen requirement, and use of opioids, diuretics and anticholinergic medication (dehydration score); vascular obstruction, and use of non-steroidal anti-inflammatory drugs (NSAIDs), steroids and diuretics (edema score); peritoneal and liver metastasis (ascites); lung and pleural metastasis, and pneumonia (pleural effusion); and intestinal obstruction and oral intake of fluids (for all symptoms) [6Go, 8Go, 15Go, 28Go].

Statistical analyses
We analyzed data for patients who died at least 3 weeks after their initial evaluation. The rationale for this decision was that we had no appropriate instruments to indicate reliable base-line points for analyses, and hydration therapy was likely to influence patient symptoms after a considerable time lag (i.e. hydration volume the patients had received 1–3 weeks before death could affect patients symptoms 48 h before death). To examine a bias, we compared patient backgrounds between the excluded and included patients.

We divided patients into two groups: those who received artificial hydration of 1 l/day or more both 1 week and 3 weeks before death (hydration group: total n=59; 31 from oncology and 28 from palliative/home-care settings) and those who did not (non-hydration group: total n=167 [18], from oncology and 149 from palliative/home-care settings). This classification was determined on the basis of actual data distributions, and the results using the other classifications achieved the similar conclusions.

To explore the potential association between hydration groups and patient symptoms, we compared the number of patients whose symptom scores increased in the final 3 weeks (dehydration and edema scores by three or more points; ascites and pleural effusion scores by one or more point) between the hydration and non-hydration groups. The results using the other cut-off points achieved the same conclusions. We also compared the prevalence of bronchial secretion, hyperactive delirium, myoclonus and bedsores, the degree of communication capacity, and the degree of agitation between the two groups.

To explore the effects of covariate factors and treatment settings, we examined the potential interactions between hydration groups and changes in dehydration score, edema score, and ascites and pleural effusion severity scores by the repeated measurement analysis with the covariates entered into the models (robust variance with the Proc mixed procedure). No covariates except for peritoneal metastasis and treatment settings statistically influenced the outcomes. In addition, subgroup analyses for patients who drank <500 ml/day of fluids throughout the last 3 weeks of life (n=108), patients with intestinal obstruction (n=114) and patients who received no intestinal drainage (n=192) achieved the same results. We therefore reported the results for the entire sample with adjusted P values to allow for difference in peritoneal metastasis and treatment settings, as well as subgroup analysis of patients with peritoneal metastases (n=145).

Finally, to provide additional information for interpreting data, we compared the changes in blood urea nitrogen/creatinine levels between hydration and non-hydration groups using repeated measurement analysis. We also calculated the prevalence of fluid retention symptoms 24 h before death among dehydrated patients, defined as presence of dry axillary (diagnosis on the basis of sunken eyes achieved similar results).

Univariate analyses were conducted using the {chi}2-test (Fisher's exact method) and the Mann–Whitney U-test, where appropriate. All analyses were performed using the statistical package SAS.


    Results
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Patient background
All 498 patients who met the inclusion criteria were consecutively recruited for this study, but a total of 272 patients were excluded for the following reasons: death within 3 weeks of initial assessment (n=200), survival beyond the observation period (n=35), medical complications (n=17), prior communication difficulty (n=15) and discharge (n=5). Thus, a total of 226 patients (49 from oncology units and 177 from palliative/home-care settings) were finally analyzed. There were no statistically significant differences in patient age and primary tumor sites between the patients excluded from the study due to death within 3 weeks and those analyzed, but the former was more likely to be male (Table 1).


View this table:
[in this window]
[in a new window]
 
Table 1. Characteristics of excluded and included patients

 
Patient backgrounds are summarized in Table 2. There were significant differences in primary tumor sites, prevalence of lung and peritoneal metastases, vascular obstruction, intestinal obstruction, the use of NSAIDs and steroids, and oral intake 3 weeks and 1 week before death between the hydration and non-hydration groups. Chemotherapy was performed in seven patients.


View this table:
[in this window]
[in a new window]
 
Table 2. Patient characteristics

 
Table 3 summarizes hydration practice in the study subjects. The mean hydration volume in the hydration group ranged from 838 to 1405 ml/day during the last 3 weeks, and the median hydration volume in the non-hydration group was 200 ml/day at all three observation points.


View this table:
[in this window]
[in a new window]
 
Table 3. Hydration practice in the final 3 weeks

 
At baseline, ascites was present but asymptomatic in 27% (n=62) and symptomatic in 20% (n=44) of all patients. Pleural effusion was present but asymptomatic in 12% (n=27) and symptomatic in 6.6% (n=15) of all patients.

Dehydration
The percentage of patients whose dehydration score increased by three or more points in the final 3 weeks of life was significantly higher in the non-hydration group than in the hydration group [35% (n=59) versus 14% (n=8); P=0.0020]. After controlling for covariates and treatment settings, there was a statistically significant interaction between hydration group and changes in the dehydration score (1.6 ± 1.4 3 weeks before death to 2.7 ± 1.6 24 h before death in the hydration group versus 1.3 ± 1.3 to 3.2 ± 1.5 in the non-hydration group; P=0.0043) (Figure 1).



View larger version (16K):
[in this window]
[in a new window]
 
Figure 1. Effects of hydration on dehydration and fluid retention symptoms. Open circles, hydration group (n=59); filled circles, non-hydration group (n=167).

 
Edema
The number of patients whose edema scores increased by three or more points was significantly higher in the hydration group than in the non-hydration group [44% (n=26) versus 29% (n=49); P=0.039]. After controlling for covariates and treatment settings, the interaction between hydration group and changes in the edema score did not reach statistical significance (2.2 ± 3.3 3 weeks before death to 6.1 ± 6.4 24 h before death in the hydration group versus 3.5 ± 4.5 to 5.2 ± 5.2 in the non-hydration group; P=0.15) (Figure 1).

Ascites
The percentage of patients whose symptom score increased by one or more point during the final 3 weeks was significantly higher in the hydration group than in the non-hydration group [29% (n=17) versus 8.4% (n=14); P <0.001]. After controlling for covariates and treatment settings, there was a statistically significant interaction between hydration group and changes in the ascites score (0.73 ± 0.78 3 weeks before death to 0.92 ± 0.88 24 h before death in the hydration group versus 0.64 ± 0.79 to 0.58 ± 0.74 in the non-hydration group; P=0.035) (Figure 1).

Pleural effusion and bronchial secretion
The number of the patients whose pleural effusion symptom score increased by one or more point in the final 3 weeks was significantly higher in the hydration group than in the non-hydration group [15% (n=9) versus 5.4% (n=9); P=0.016]. Hydration group was not significantly associated with changes in the pleural effusion symptom score after controlling for covariates and treatment settings (0.22 ± 0.46 3 weeks before death to 0.36 ± 0.61 24 h before death in the hydration group versus 0.27 ± 0.60 to 0.31 ± 0.63 in the non-hydration group; P=0.76) (Figure 1).

There was no statistically significant difference in the prevalence of bronchial secretion between the hydration and the non-hydration groups (Table 4).


View this table:
[in this window]
[in a new window]
 
Table 4. Symptom severity in the last 3 weeks of the patients with and without hydration

 
Communication capacity, agitation and delirium
There were no statistically significant differences in the communication score, agitation score or prevalence of hyperactive delirium between the hydration and the non-hydration groups (Table 4).

Myoclonus and bedsores
There were no statistically significant differences in the prevalences of myoclonus or bedsores between the hydration and the non-hydration groups [for myoclonus 1.7% (n=1) versus 8.4% (n=14), P=0.12; for bedsores 27% (n=16) versus 34% (n=57), P=0.32].

Patients with peritoneal metastases
The interactions between hydration group and symptom changes in the last 3 weeks were statistically significant in dehydration score (1.7 ± 1.43 weeks before death to 2.9 ± 1.6 24 h before death in the hydration group versus 1.3 ± 1.3 to 3.5 ± 1.4 in the non-hydration group; P=0.0043) and pleural effusion score (0.22 ± 0.47 to 0.35 ± 0.60 versus 0.30 ± 0.63 to 0.27 ± 0.57, respectively; P=0.046), and marginally significant in ascites score (0.91 ± 0.78 to 1.0 ± 0.87 versus 0.88 ± 0.80 to 0.70 ± 0.75, respectively; P=0.091).

Laboratory findings
We obtained paired blood samples taken 3 weeks and 1 week before death from 37 (63%) and 56 (34%) patients in the hydration and non-hydration groups, respectively. The blood urea nitrogen/creatinine levels increased from 34 ± 15 to 44 ± 18 mg/dl in the hydration group in the last 3 weeks, compared with from 31 ± 17 to 39 ± 20 mg/dl in the non-hydration group. The difference between hydration groups was not statistically significant (P=0.58).

Comorbidity of dehydration and fluid retention symptoms
Of the 149 dehydrated patients with dry axillary 24 h before death, 73%, 46% and 19% had simultaneous edema, ascites or pleural effusion, respectively; and 81% had some fluid retention symptoms (Table 5).


View this table:
[in this window]
[in a new window]
 
Table 5. Fluid retention symptoms in dehydrated patients (n=149)

 

    Discussion
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
This is, to the best of our knowledge, the largest and the first multicenter observation study to investigate the association between hydration volume and dehydration and fluid retention symptoms in terminally ill cancer patients.

This study revealed that peripheral edema, ascites and pleural effusion in the hydration group were more likely to worsen in the last 3 weeks. The association between hydration group and ascites severity was statistically significant after controlling all covariates and treatment settings, and in a subgroup of patients with peritoneal metastases there was a statistically significant interaction between hydration practice and changes in pleural effusion severity. The underlying mechanisms of fluid retention symptoms include a decrease in colloid osmotic pressure, an increase in membrane permeability, and an increase in hydrostatic pressure [11Go]. Our findings suggest that overhydration in the terminal phase could deteriorate fluid retention symptoms.

We also found that dehydration scores increased in the last 3 weeks of life regardless of whether patients received artificial hydration or not, although scores increased less in the hydration than in the non-hydration group. The potential interpretations of this finding are that: (i) the instruments for measurement of dehydration used in this study could not differentiate dehydration signs from changes related to progressed cachexia; (ii) current hydration volume was not sufficient to maintain hydration status and more active hydration could alleviate membrane dehydration signs; or (iii) artificial hydration therapy in the terminal stage could not effectively alleviate dehydration even if an appropriate volume was provided due to some pathological mechanisms (e.g. fluid shift from the intravascular components to the third space). The first interpretation we consider unlikely, because additional analysis of laboratory findings suggested that the blood urea nitrogen/creatinine level, a biological marker of dehydration, increased in the hydration group similar to the non-hydration group. The second interpretation seems also less likely, because a well-conducted open trial and a small randomized controlled trial indicated that artificial hydration therapy had limited beneficial effects in alleviation of thirst sensation for most terminally ill cancer patients [7Go, 9Go], and this study suggests that more active hydration would cause more fluid retention symptoms, limiting the use of hydration. On the other hand, the third interpretation is supported by an exploratory study indicating that the main pathophysiology of terminal dehydration is decreased intravenous volume with increased interstitial fluids [11Go], and this study revealed that many patients simultaneously had both dehydration and fluid retention symptoms. Therefore, it is suggested that while artificial hydration therapy may help alleviate membranous dehydration signs in some patients, the overall benefits of active hydration therapy are limited by the possibility of aggravating fluid retention symptoms [14Go, 15Go].

This study did not identify any beneficial effects of artificial hydration therapy on psychiatric symptoms. Previous retrospective, historical control and prospective observational studies have demonstrated that active rehydration could contribute to alleviation of delirium [12Go, 13Go, 16Go], while another historical control study and a small randomized controlled trial found no overall benefit [7Go, 27Go]. These conflicting results suggest that the benefits of artificial hydration therapy in alleviating delirium may be applied to a certain group of patients with specific underlying etiologies, such as opioid hyperexcitability syndrome or acute dehydration [5Go].

This study identified no clear association between hydration volume and the development of bronchial secretion. Of note was that our sample was limited to patients with abdominal malignancies, and hydration volume was relatively small. Therefore, our findings suggested that, for patients with abdominal malignancies receiving moderate level of hydration (e.g. ≤1 l/day), bronchial secretion is not influenced by hydration volume. On the other hand, previous observational studies including lung cancer patients have identified pulmonary edema as a significant etiology of severe bronchial secretion [10Go, 29Go], and bronchial secretion has multiple etiologies, including respiratory malignancies, infection, pulmonary edema, dysphasia and brain metastases [30Go, 31Go]. Thus, the effect of hydration volume on other groups of patients should be examined in future studies.

This study successfully recruited patients with a narrow range of primary tumor sites, enrolled patients from multiple centers, used a comprehensive set of assessments that were sensitive to symptom changes and highly feasible, and prospectively evaluated multiple symptoms. Nonetheless, this study has several limitations. First, this was not an intervention trial. Although we acknowledge that a randomized controlled study is the best research design to scientifically clarify the treatment effects of hydration therapy, the information required for planning controlled trials, such as useful end point measures, their estimated differences and the necessary sample size, is lacking. Therefore, we decided to perform an observation study first. Secondly, the main end points were measured objectively. Therefore, we did not evaluate the effect of hydration volume on patients' subjective well-being, and there was a possibility of under- or overestimation in addition to reporting bias from treating physicians. This is, we believe, a realistic option to minimize selection bias and ensure sufficient sample size, but this flaw should be overcome in the next study. Future studies should adopt a combination of patient-rated well-being and the objective methods successfully used in this study as the primary end points. Thirdly, the reliability and validity of some measurements (i.e. peripheral edema, ascites and pleural effusion) have not been formally tested. We minimized this potential bias by confirming the full agreement of physicians and nurses, and explicitly defining the criteria in rating systems. Fourthly, stomach cancer is one of the most common malignancies in Japan, and was the primary diagnosis in nearly 30% of our subjects. Our findings therefore may not be generalizable to patients from other countries. Fifthly, as only patients who eventually died were analyzed, we did not evaluate the effects of hydration on patient survival. Finally, the result could be influenced by the treatment bias: it is possible that dehydration symptoms in the non-hydration group would have improved if they had received hydration, or that fluid retention symptoms in the hydration group would have been minimized if they had not received hydration.

In conclusion, although artificial hydration therapy might alleviate membranous dehydration signs in terminally ill patients, it could worsen peripheral edema, ascites and pleural effusions. Our findings suggest that the potential benefits of artificial hydration therapy should be balanced with the risk of worsening fluid retention symptoms. Further clinical studies are clearly needed to identify which subgroups of terminally ill patients may or may not benefit from artificial hydration therapy. In the meantime, an individualized treatment based on the comprehensive assessment followed by close monitoring of both dehydration and fluid retention symptoms is strongly recommended.


    Acknowledgements
 
We would like to thank Takuhiro Yamaguchi, PhD (Department of Biostatistics, The University of Tokyo) for advice on statistical analyses. This work was supported in part by a Grant-in-Aid for Cancer Research (11-2) from the Ministry of Health and Welfare, Japan. The members of the Japan Palliative Oncology Study Group are: Taketoshi Ozawa (Palliative Care Unit, Yokohama Kosei Hospital); Chikako Kato (Hospice, St Francis Hospital); Yoshiyuki Kizawa (Palliative Care, Tsukuba Medical Center Hospital); Takeshi Okabe (Okabe Clinic); Hiroyuki Kohara (Palliative Care, National Sanyo Hospital); Yasuo Shima (Palliative Care, National Cancer Hospital East); Kayako Shimamura (Hospice, Kyuseigun Kiyose Hospital); Yoshikazu Chinone (Palliative Care, Japanese Red Cross Medical Center); Yoshi Tsukiyama (Department of Palliative Medicine, Wakayama Medical University); Junichi Tsunoda (Palliative Care, Sannou Hospital); Kenji Nishitateno (General and Medical Director, Peace House Hospice); Akitoshi Hayashi (Director of Hospice Service, The Japan Baptist Hospital); Yoshifumi Honke (Department of Palliative Care, Hiroshima Prefectural Hospital); Hideaki Nagai (Department of Palliative Medicine, National Tokyo Hospital); Makoto Miyoshi (Palliative Care, Kitakyushu Municipal Medical Center); Kinomi Shinomiya (Palliative Care, Saitama Cancer Center); Fumiki Asanuma (Department of Surgery, Kitasato Institute Hospital); Yoshikazu Ashino (Department of Surgery, Fukushima-rosai Hospital); Hisanao Ohkura (Medical Oncology Division, Ibaraki Kenritsu Chuo Hospital & Cancer Center); Hiroyuki Otani (Department of Internal Medicine, Second Teaching Hospital, Fujita Health University School of Medicine); Nobuhiro Tsukada (Department of Internal Medicine, Saiseikai Center Hospital, Tokyo); Nobuhisa Nakajima (Department of Surgery, Sapporo Social Insurance General Hospital); Akira Hoshino (Department of Surgery, Iwate Prefectural Kitakami Hospital); Toshihiro Matsuura (Department of Gastroenterology, Chubu National Hospital); Takatsugu Kawagoe (Department of Surgery, Toki General Hospital); Takuya Aoki (Respiratory Division, Department of Internal Medicine, Saiseikai Central Hospital); and Takayuki Kaburagi (Internal Medicine, Ibaraki Kenritsu Chuo Hospital & Cancer Center).


    Notes
 
{dagger} Members of the Japan Palliative Oncology Study Group are listed in the Acknowledgements. Back

Received for publication August 13, 2004. Revision received October 25, 2004. Accepted for publication November 29, 2004.


    References
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
1. Sarhill N, Walsh D, Nelson K, Davis M. Evaluation and treatment of cancer-related fluid deficits: volume depletion and dehydration. Support Care Cancer 2001; 9: 408–419.[CrossRef][ISI][Medline]

2. Bruera E, Belzile M, Watanabe S, Fainsinger RL. Volume of hydration in terminal cancer patients. Support Care Cancer 1996; 4: 147–150.[ISI][Medline]

3. Morita T, Shima Y, Adachi I, Japan Palliative Oncology Study (J-POS) Group. Attitudes toward terminal dehydration of Japanese physicians: a nationwide survey. J Clin Oncol 2002; 20: 4699–4704.[Abstract/Free Full Text]

4. Andrews M, Bell ER, Smith SA et al. Dehydration in terminally ill patients. Is it appropriate palliative care? Postgrad Med J 1993; 93: 201–208.[Medline]

5. Lawlor PG. Delirium and dehydration: some fluid for thought? Support Care Cancer 2002; 10: 445–454.[CrossRef][ISI][Medline]

6. Burge FI. Dehydration symptoms of palliative care cancer patients. J Pain Symptom Manage 1993; 8: 454–464.[CrossRef][ISI][Medline]

7. Cerchietti L, Navigante A, Sauri A, Palazzo F. Hypodermoclysis for control of dehydration in terminal-stage cancer. Int J Palliat Nurs 2000; 6: 370–374.[Medline]

8. Ellershaw JE, Sutcliffe JM, Saunders CM. Dehydration and the dying patient. J Pain Symptom Manage 1995; 10: 192–197.[CrossRef][ISI][Medline]

9. McCann RM, Hall WJ, Grith-Juncker A. Comfort care for terminally ill patients. The appropriate use of nutrition and hydration. JAMA 1994; 272: 1263–1266.[Abstract]

10. Morita T, Tsunoda J, Inoue S, Chihara S. Risk factors for death rattle in terminally ill cancer patients. A prospective exploratory study. Palliat Med 2002; 14: 19–23.[CrossRef]

11. Morita T, Tei Y, Inoue S et al. Fluid status of terminally ill cancer patients with intestinal obstruction: an exploratory observational study. Support Care Cancer 2002; 10: 474–479.[CrossRef][ISI][Medline]

12. Bruera E, Franco J, Maltoni M et al. Changing pattern of agitated impaired mental status in patients with advanced cancer: association with cognitive monitoring, hydration, and opioid rotation. J Pain Symptom Manage 1995; 10: 287–291.[CrossRef][ISI][Medline]

13. Lawlor PG, Ganon B, Mancini IL et al. Occurrences, causes, and outcome of delirium in patients with advanced cancer. A prospective study. Arch Intern Med 2000; 160: 786–794.[Abstract/Free Full Text]

14. Mercadante S, Ripamonti C, Casuccio A et al. Comparison of octreotide and hyoscine butylbromide in controlling gastrointestinal symptoms due to malignant inoperable bowel obstruction. Support Care Cancer 2000; 8: 188–191.[CrossRef][ISI][Medline]

15. Morita T, Tsunoda J, Inoue S, Chihara S. Determinants of sensation of thirst in terminally ill cancer patients. Support Care Cancer 2001; 9: 177–186.[CrossRef][ISI][Medline]

16. de Stoutz ND, Bruera E, Suarez-Almazor M. Opioid rotation for toxicity reduction in terminal cancer patients. J Pain Symptom Manage 1995; 10: 378–384.[CrossRef][ISI][Medline]

17. Viola RA, Wells GA, Peterson J. The effects of fluid status and fluid therapy on the dying: a systemic review. J Palliat Care 1997; 13: 41–52.[ISI][Medline]

18. Morita T, Adachi I, Japan Palliative Oncology Study (J-POS) Group. Satisfaction with rehydration therapy for terminally ill cancer patients: concept construction, scale development, and identification of contributing factors. Support Care Cancer 2002; 10: 44–50.[CrossRef][ISI][Medline]

19. Pereira J, Hanson J, Bruera E. The frequency and clinical course of cognitive impairment in patients with terminal cancer. Cancer 1997; 79: 835–842.[CrossRef][ISI][Medline]

20. Stromgren AS, Goldschmidt D, Groenvold M et al. Self-assessment in cancer patients referred to palliative care. A study of feasibility and symptom epidemiology. Cancer 2002; 94: 512–520.[CrossRef][ISI][Medline]

21. Eaton D, Bannister P, Mulley GP, Connolly M. Auxiliary sweating in clinical assessment of dehydration in ill elderly patients. BMJ 1994; 308: 1271.[Free Full Text]

22. Gross CR, Lindquist RD, Woolley AC et al. Clinical indicators of dehydration severity in elderly patients. J Emerg Med 1992; 10: 267–274.[Medline]

23. McGee S, Abernethy WB, Simel DL. Is this patient hypovolemic? JAMA 1999; 281: 1022–1029.[Abstract/Free Full Text]

24. Back IN, Jenkins K, Blower A, Beckhelling J. A study comparing hyoscine hydrobromide and glycopyrrolate in the treatment of death rattle. Palliat Med 2001; 15: 329–336.[CrossRef][ISI][Medline]

25. Morita T, Tusnoda J, Inoue S et al. Communication Capacity Scale and Agitation Distress Scale to measure the severity of delirium in terminally ill cancer patients: a validation study. Palliat Med 2001; 15: 197–206.[CrossRef][ISI][Medline]

26. Matsuoka Y, Miyake Y, Arakaki H et al. Clinical utility and validation of the Japanese version of Memorial Delirium Assessment Scale in a psychogeriatric inpatient setting. Gen Hosp Psychiatry 2001; 23: 36–40.[CrossRef][ISI][Medline]

27. Morita T, Tei Y, Inoue S. Agitated terminal delirium and association with partial opioid substitution and hydration. J Palliat Med 2003; 6: 557–563.[CrossRef][Medline]

28. Morita T, Tsunoda J, Inoue S, Chihara S. Contributing factors to physical symptoms in terminally-ill cancer patients. J Pain Symptom Manage 1999; 18: 338–346.[CrossRef][ISI][Medline]

29. Wildiers H, Menten J. Death rattle: prevalence, prevention and treatment. J Pain Symptom Manage 2002; 23: 310–317.[CrossRef][ISI][Medline]

30. Bennett M, Lucas V, Brennan M et al. Using anti-muscarinic drugs in the management of death rattle: evidence-based guidelines for palliative care. Palliat Med 2002; 16: 369–374.[CrossRef][ISI][Medline]

31. Morita T, Hyodo I, Yoshimi T et al. Incidence and underlying etiologies of bronchial secretion in terminally ill cancer patients: a multi-center prospective observation study. J Pain Symptom Manage 2004; 27: 533–539.[CrossRef][ISI][Medline]





This Article
Abstract
Full Text (PDF)
All Versions of this Article:
16/4/640    most recent
mdi121v1
E-letters: Submit a response
Alert me when this article is cited
Alert me when E-letters are posted
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Search for citing articles in:
ISI Web of Science (1)
Disclaimer
Request Permissions
Google Scholar
Articles by Morita, T.
PubMed
PubMed Citation
Articles by Morita, T.