Anaesthesia 2014, 69, 1127–1132
doi:10.1111/anae.12756
Original Article Psychological outcomes in critically ill patients receiving renal replacement therapy: a matched-pairs analysis* J. J. Paddle,1 K. A. Sharpe,2 N. Boyd2 and K. Burt3 1 Consultant in Intensive Care Medicine, 2 Specialist Registrar in Anaesthesia, 3 Research Nurse, Intensive Care Unit, Royal Cornwall Hospital NHS Trust, Truro, UK
Summary We conducted a single-centre observational study over five years to assess the impact of renal replacement therapy on the psychological health of survivors of critical illness. We hypothesised that the added burden of renal replacement would increase the prevalence and severity of anxiety, depression and stress reactions in these patients, compared with matched pairs (matched for age, sex and APACHE II score) who did not receive renal replacement. Participants completed postal questionnaires. A total of 342 patients with acute kidney injury received renal replacement. One hundred and seventy-nine (52.3%) survived to hospital discharge, and 161 (47.1%) were alive at 90 days. Seventy-seven (47.8% of survivors) completed questionnaires. We found 77 matches for the Hospital Anxiety and Depression Scale analysis and 72 for the Impact of Events Scale analysis. Clinically relevant symptoms of psychiatric morbidity were common, with anxiety and depression affecting 49 (63.6%) patients and stress reactions affecting 24 (33.3%) patients. Mean scores (95% CI) were 11.4 (9.6–13.2) and 20.1 (15.7–24.6), respectively. On multivariate analysis, we found no significant differences between renal replacement patients and controls, in either the frequency or severity of these symptoms. .................................................................................................................................................................
Correspondence to: J. J. Paddle Email: [email protected] *Presented in part at the 26th Annual Congress of the European Society of Intensive Care Medicine, Paris, France; October 2013. Accepted: 26 April 2014
Introduction Acute kidney injury (AKI) requiring renal replacement therapy (RRT) is common in critically ill patients, with an incidence of up to 9.4% [1]. It is associated with a poor prognosis, with a hospital mortality of between 35% and 63.8% [2, 3]. Among survivors, there is a continued adverse impact on quality of life (QoL) [4]. Studies that have assessed the impact of AKI with or without RRT on QoL have used global measures that incorporate physical, pain, social and other parameters, as well as psychological morbidity [2, 5–12]. There is a paucity of data on the specific psychological impact of RRT. © 2014 The Association of Anaesthetists of Great Britain and Ireland
We sought to assess the impact of continuous RRT (CRRT) on the psychological health of survivors of critical illness. Our hypothesis was that the added burden of CRRT would increase the prevalence and severity of anxiety, depression and stress reactions in this group of patients, compared with a matched group of critical illness survivors who did not require CRRT.
Methods We conducted an observational study of survivors of critical illness who had received CRRT in our institution, and compared them with matched pairs of 1127
Anaesthesia 2014, 69, 1127–1132
survivors of critical illness who had not received CRRT. The study was approved by the Cornwall and Plymouth Research Ethics Committee, and performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki. As the study was purely observational, and used existing, routinely collected data, informed consent was not required from the subjects. We collected data on all patients receiving CRRT in our intensive care unit (ICU), which is a mixed adult medical and surgical general ICU, over a fiveyear period from 1 January 2007 to 31 December 2011. All patients receiving CRRT met the RIFLE (Risk, Injury, Failure, Loss and End-stage kidney disease) criteria for AKI. No patient was treated with intermittent haemodialysis. Patients who survived to three months after hospital discharge were sent a standardised QoL postal questionnaire incorporating the Hospital Anxiety and Depression Scale (HADS) and Impact of Events Scale (IES). We sought matched controls from critically ill patients on our unit who had not received CRRT. Each patient receiving CRRT was matched with a control patient according to the following criteria: admission to the critical care unit within one year of index case; same sex; age within five years; and APACHE II score within five points. This was defined as a good match. Where more than one match was identified, a random number generator (from within Microsoft Excel software) was used to select a single match. Where no match was found, the matching criteria were expanded to a control patient within 10 years of age, and within 10 APACHE II points. This was defined as a poor match. All critically ill patients in our institution are sent a postal questionnaire three months after hospital discharge, which incorporates the HADS and IES, and we used this to evaluate aspects of psychological morbidity. We used routinely collected data from our inpatient database for all demographic and survival data. We conducted multivariate analysis of matched pairs using a general linear model to compare HADS and IES between CRRT patients and controls. We used the Statistical Package for the Social Sciences software version 19 (SPSS Inc., Chicago, IL, USA) for this and for all descriptive statistics. Two-by-two chi-squared 1128
Paddle et al. | Psychological outcomes of dialysis in critical care
tests were used to compare numbers of patients with scores consistent with clinical anxiety, depression and impact of events. Statistical significance was assumed at p < 0.05.
Results During the study period, 395 patients received CRRT on our critical care unit, of which 342 met the RIFLE criteria for AKI. Of these, 179 (52.3%) survived to hospital discharge and 161 (47.1%) were alive at 90 days. All were sent questionnaires, and 77 (47.8% of survivors at 90 days) returned these completed. Figure 1 shows the CONSORT flow diagram for the study. Matches for these 77 patients were sought from our ICU database. Table 1 shows the baseline characteristics of the CRRT group and controls.
Enrollment
Assessed for eligibility (n = 395) Excluded • Not meeting RIFLE criteria for AKI (n = 53) Enrolled (n = 342) Died • Hospital death (n = 163) • Post-discharge, before 90 days (n = 18)
Survived (n = 161) Postal questionnaires sent to all No reply to questionnaire (n = 84)
Allocation
Database of ICU survivors with completed questionnaires (n = 338)
Analysis
Completed questionnaires (n = 77)
Matched pairs found for HADS analysis (n = 77)
Matched pairs found for IES analysis (n = 72)
Figure 1 CONSORT flow diagram. RIFLE, Risk, Injury, Failure, Loss and End-stage kidney disease; AKI, acute kidney injury; HADS, Hospital Anxiety and Depression Scale; IES, Impact of Events Scale. © 2014 The Association of Anaesthetists of Great Britain and Ireland
Paddle et al. | Psychological outcomes of dialysis in critical care
Table 1 Characteristics of patients who underwent continuous renal replacement therapy (CRRT) and control patients. Values are mean (95% CI) or number (proportion).
All matches n Age; years APACHE II score Male Good matches only n Age APACHE II score Male
CRRT
Control
77 63 (60–66) 24.3 (23.1–25.4) 45 (58.4%)
77 64 (61–67) 22.5 (21.4–23.6) 45 (58.4%)
65 66 (63–69) 24.2 (22.9–25.5) 40 (61.5%)
65 66 (63–69) 23.0 (21.8–24.1) 40 (61.5%)
p value
0.037 < 0.001
1.00
Anaesthesia 2014, 69, 1127–1132
Table 2 Hospital Anxiety and Depression Scale (HADS) and Impact of Events Scale (IES) in patients who underwent continuous renal replacement therapy (CRRT) and control patients. Values are mean (95% CI) or number (proportion).
HADS n Score Patients scoring ≥ 8 IES n Score Patients scoring ≥ 24
CRRT
Control
p value
77 11.4 [9.6–13.2] 49 [63.6%]
77 11.7 [9.7–13.7] 49 [63.6%]
0.846 1.00
72 20.1 [15.7–24.6] 24 [33.3%]
72 21.9 [16.2–26.2] 26 [36.1%]
0.752 0.86
< 0.001
The two groups were broadly comparable in terms of age and sex match. However, the APACHE II scores were significantly higher in the CRRT group. Although the difference was numerically smaller when only good matches were compared, a significant difference remained. This reflected the matching process that allowed for absolute differences in APACHE II score (as it would not have been possible to find matches with exactly the same scores in the database), and tended to select control patients that were slightly less sick. The multivariate analysis allowed for the assessment of the interaction of these variables within the HADS and IES scores where no significant interactions were found. Seventy-seven CRRT patients completed the questions required to calculate a HADS score, as did all 77 matched controls. Two of the CRRT patients did not complete sufficient questions to calculate an IES score, nor did three (different) control patients. The IES analysis was therefore conducted on 72 matched pairs. Table 2 illustrates the mean scores for the two tests, comparing CRRT patients with controls, and the number of patients with clinically relevant symptoms of anxiety, depression and post-traumatic stress reactions. We used a HADS score of 8 or more and an IES score of 24 or more as thresholds of clinical relevance [13, 14], and found no differences between the two groups for either test. © 2014 The Association of Anaesthetists of Great Britain and Ireland
Using a general linear model and multivariate analysis, accounting for any differences between the two groups with respect to age or APACHE II score, we found no significant differences between the HADS scores for CRRT patients and the matched controls. This remained the case whether we included all matches, or restricted the analysis to only the 65 good matches. Similarly, and using the same model, we found no significant differences between the IES scores in the two groups.
Discussion To the best of our knowledge, this is the first study that specifically assesses the impact of CRRT on the psychological health of critically ill patients in comparison with matched control patients who do not receive CRRT. Symptoms of anxiety and depression were common in our patients, affecting about two thirds of both patients receiving CRRT and matched controls. About a third of patients in both groups exhibited symptoms of clinically relevant post-traumatic stress. Similar high levels of psychiatric morbidity in critically ill patients have been found by other authors [15]. Among patients receiving haemodialysis for end-stage renal disease, symptoms of anxiety and depression are common, with 45.7% of subjects meeting criteria for an anxiety disorder [16] and 23.7% with significant depressive symptoms [17]. Symptoms consistent with post-traumatic stress are also common in this patient 1129
Anaesthesia 2014, 69, 1127–1132
group, with 77.8% of patients reporting at least one traumatic event [18]. Other studies have used global measures of QoL in this patient population, that incorporate domains pertinent to mental health. The Finnish Acute Kidney Injury study [11] assessed QoL in survivors of critical illness with AKI across ten centres in Finland, using the EuroQol-5D (EQ-5D) questionnaire. Subjects were assessed at six months. The EQ-5D assesses QoL across five health dimensions, of mobility, self-care, usual activities, pain or discomfort and anxiety or depression. The anxiety and depression domain asks only three questions and there is no assessment of potential impact of events on mental health. Another Finnish study [12] did look specifically at patients receiving RRT, but again used the EQ-5D assessment. Both studies found that QoL was poor by objective assessment in survivors of AKI, but that patients rated their subjective health to be as good as matched controls from the general population. Two European studies used the Short Form Health Survey (SF-36) questionnaire to report QoL in ICU patients with AKI [8, 9]. This assessment is also of global measures of health, covering eight health dimensions, of which mental health is just one, and the scores are aggregated. Again, no assessment is made of the potential impact of events. One study [8] reported a reduction in QoL at six months, the other [9], which repeated the assessment at different time intervals, reported a reduction in three domains of the SF-36 for ICU patients with AKI compared with those without at ICU discharge, but this difference was no longer present at six months. The Nottingham health profile, or simplified versions of it, has been used to assess QoL in three studies from Australia [5] and Europe [6, 7]. All were restricted to patients from ICUs receiving RRT. Two studies assessed QoL at six months; one study did not report the time interval. Again, this assessment measures a global QoL based on activity and functional health and is not specific to psychological health. All three studies reported that survivors had a good or acceptable QoL. We did not find any difference in measures of psychiatric morbidity between patients receiving CRRT and controls, and this remained the case for both 1130
Paddle et al. | Psychological outcomes of dialysis in critical care
simple chi-squared analysis and multivariate analysis that accounted for baseline differences. We therefore believe this study provides evidence that the added burden of RRT does not adversely affect the psychological health of critically ill patients. There may be a number of reasons for this. Firstly, AKI that is sufficiently severe to require RRT generally occurs during, and is part of, the critical illness, such that the patient is overwhelmed by stressors to the extent that RRT adds little to the perceived stress of the total illness period. Secondly, many of these patients will be sedated during their critical illness in order to facilitate treatment modalities such as mechanical ventilation, so the patient may not even be aware that he or she is receiving RRT. In our ICU, sedation is achieved with a combination of propofol and opioid infusion, titrated to a defined sedation target, and this applies both to patients who receive CRRT and to those who do not. Our findings are supported by the small number of other studies that have compared QoL outcomes in ICU patients with and without AKI [2, 9, 11]. None of these studies looked specifically at psychological morbidity, and only one [11] included a matched-pairs analysis. Even if RRT does increase psychological morbidity, it may be that this does not persist when the AKI resolves. While we did not specifically assess long-term dependence on RRT for this study, we know that in our own cohort of critically ill patients who receive RRT, only 3.7% still require RRT at 30 days [19], similar to an international finding of 5.3% of patients requiring RRT at hospital discharge [3]. Thus it may be that by three months after illness, any morbidity attributable to the period of RRT during critical illness has resolved for the overwhelming majority of patients who are no longer dialysis dependent. Conversely, it is possible that there is a difference in psychological morbidity between the two groups that our study failed to detect. Our study may have been under-powered, even though we sought to include all survivors over a five-year period. We chose assessment tools that have been validated in various patient groups [13, 14], including patient with endstage renal disease receiving haemodialysis or peritoneal dialysis [20] such that we would expect the HADS and IES scores to provide a better assessment of psychological trauma than a more generic score of QoL. © 2014 The Association of Anaesthetists of Great Britain and Ireland
Paddle et al. | Psychological outcomes of dialysis in critical care
The response rate to our questionnaires was relatively low, so there is a chance of selection bias. Patients suffering from psychiatric morbidity might be more inclined to respond as a means of seeking help. Alternatively, survivors with psychological sequelae might be less likely to respond, owing to the emotional distress that answering the questions might bring. We cannot exclude this potential bias. However, the response rate in this study is similar to the response rate we see for all ICU survivors (51.8%, unpublished clinic data), such that any bias would be expected to be matched in the control group. The timing of the survey may have influenced the responses, and some authors have observed that QoL improves over time following the illness [4]; however, that improvement may be restricted to physical and emotional domains of QoL, and not to mental health [4]. Other authors have failed to find any association between timing of survey and QoL [12]. In any event, both RRT and control groups were surveyed at the same three-month interval after discharge. Other limitations of our study should be noted. As it is a retrospective observational study, we are not able to comment on causality. As a single-centre study, our findings may not be applicable to other geographical areas or ethnic groups, however, our patients’ baseline demographics are similar to those of other critically ill patients in general hospitals across the UK. In conclusion, our five-year observational study of survivors of critical illness demonstrated that psychiatric morbidity, as assessed by HADS and IES, is common. However, our analysis found that in an ICU population matched for age, sex, and severity of illness, there was no additional psychological morbidity from RRT.
Acknowledgements The authors thank Jo Palmer from the Research and Development department of the Royal Cornwall Hospital for her assistance with statistical analysis.
Competing interests No external funding and no competing interests declared. © 2014 The Association of Anaesthetists of Great Britain and Ireland
Anaesthesia 2014, 69, 1127–1132
References 1. Nisula S, Kaukonen K-M, Vaara ST, et al. Incidence, risk factors and 90-day mortality of patients with acute kidney injury in Finnish intensive care units: the FINNAKI study. Intensive Care Medicine 2013; 39: 420–8. 2. Vaara ST, Pettil€a V, Reinikainen M, Kaukonen K-M. Populationbased incidence, mortality and quality of life in critically ill patients treated with renal replacement therapy: a nationwide retrospective cohort study in finnish intensive care units. Critical Care 2012; 16: R13. 3. Uchino S, Bellomo R, Morimatsu H, et al. Continuous renal replacement therapy: a worldwide practice survey. Intensive Care Medicine 2007; 33: 1563–70. 4. Dowdy DW, Eid MP, Sedrakyan A, et al. Quality of life in adult survivors of critical illness: a systematic review of the literature. Intensive Care Medicine 2005; 31: 611–20. 5. Gopal I, Bhonagiri S, Ronco C, Bellomo R. Out of hospital outcome and quality of life in survivors of combined acute multiple organ and renal failure treated with continuous venovenous hemofiltration/hemodiafiltration. Intensive Care Medicine 1997; 23: 766–72. 6. Korkeila M, Ruokonen E, Takala J. Costs of care, long-term prognosis and quality of life in patients requiring renal replacement therapy during intensive care. Intensive Care Medicine 2000; 26: 1824–31. 7. Morgera S, Kraft AK, Siebert G, Luft FC, Neumayer H-H. Longterm outcomes in acute renal failure patients treated with continuous renal replacement therapies. American Journal of Kidney Diseases 2002; 40: 275–9. 8. Delannoy B, Floccard B, Thiolliere F, et al. Six-month outcome in acute kidney injury requiring renal replacement therapy in the ICU: a multicentre prospective study. Intensive Care Medicine 2009; 35: 1907–15. 9. Hofhuis JGM, van Stel HF, Schrijvers AJP, Rommes JH, Spronk PE. The effect of acute kidney injury on long-term healthrelated quality of life: a prospective follow up study. Critical Care 2013; 17: R17. 10. Johansen KL, Smith MW, Unruh ML, Siroka AM, O’Connor TZ, Palevsky PM. Predictors of health utility among 60-day survivors of acute kidney injury in the veterans affairs/national institutes of health acute renal failure trial network study. Clinical Journal of the American Society of Nephrology 2010; 5: 1366–72. 11. Nisula S, Vaara ST, Kaukonen K-M, et al. Six-month survival and quality of life of intensive care patients with acute kidney injury. Critical Care 2013; 17: R250. €m A, Tallgren M, Peltonen S, R€as€anen P, Pettil€ 12. Ahlstro a V. Survival and quality of life of patients requiring acute renal replacement therapy. Intensive Care Medicine 2005; 31: 1222–8. 13. Bjelland I, Dahl AA, Haug TT, Neckelmann D. The validity of the hospital anxiety and depression scale. An updated literature review. Journal of Psychosomatic Research 2002; 52: 69–77. 14. Asukai N, Kato H, Kawamura N, et al. Reliability and validity of the Japanese-language version of the impact of event scalerevised (IES-R-J): four studies of different traumatic events. Journal of Nervous and Mental Disease 2002; 190: 175–82. 15. Jackson JC, Hart RP, Gordon SM, Hopkins RO, Girard TD, Ely WE. Post-traumatic stress disorder and post-traumatic stress symptoms following critical illness in medical intensive care unit patients: assessing the magnitude of the problem. Critical Care 2007; 11: R27.
1131
Anaesthesia 2014, 69, 1127–1132 16. Cukor D, Coplan J, Brown C, et al. Anxiety disorders in adults treated by hemodialysis: a single-center study. American Journal of Kidney Diseases 2008; 52: 128–36. 17. Agganis BT, Weiner DE, Giang LM, et al. Depression and cognitive function in maintenance hemodialysis patients. American Journal of Kidney Diseases 2010; 56: 704–12. 18. Tagay S, Kribben A, Hohenstein A, Mewes R, Senf W. Posttraumatic stress disorder in hemodialysis patients. American Journal of Kidney Diseases 2007; 50: 594–601.
1132
Paddle et al. | Psychological outcomes of dialysis in critical care 19. Hayes-Bradley C, Caddel S, Paddle J. Observational cohort of renal replacement therapy patients at a district general hospital ICU: case mix and outcome. Critical Care 2009; 13: P272. 20. Loosman WL, Siegert CE, Korzec A, Honig A. Validity of the hospital anxiety and depression scale and the Beck depression inventory for use in end-stage renal disease patients. British Journal of Clinical Psychology 2010; 49: 507–16.
© 2014 The Association of Anaesthetists of Great Britain and Ireland
doi:10.1111/anae.12756
Original Article Psychological outcomes in critically ill patients receiving renal replacement therapy: a matched-pairs analysis* J. J. Paddle,1 K. A. Sharpe,2 N. Boyd2 and K. Burt3 1 Consultant in Intensive Care Medicine, 2 Specialist Registrar in Anaesthesia, 3 Research Nurse, Intensive Care Unit, Royal Cornwall Hospital NHS Trust, Truro, UK
Summary We conducted a single-centre observational study over five years to assess the impact of renal replacement therapy on the psychological health of survivors of critical illness. We hypothesised that the added burden of renal replacement would increase the prevalence and severity of anxiety, depression and stress reactions in these patients, compared with matched pairs (matched for age, sex and APACHE II score) who did not receive renal replacement. Participants completed postal questionnaires. A total of 342 patients with acute kidney injury received renal replacement. One hundred and seventy-nine (52.3%) survived to hospital discharge, and 161 (47.1%) were alive at 90 days. Seventy-seven (47.8% of survivors) completed questionnaires. We found 77 matches for the Hospital Anxiety and Depression Scale analysis and 72 for the Impact of Events Scale analysis. Clinically relevant symptoms of psychiatric morbidity were common, with anxiety and depression affecting 49 (63.6%) patients and stress reactions affecting 24 (33.3%) patients. Mean scores (95% CI) were 11.4 (9.6–13.2) and 20.1 (15.7–24.6), respectively. On multivariate analysis, we found no significant differences between renal replacement patients and controls, in either the frequency or severity of these symptoms. .................................................................................................................................................................
Correspondence to: J. J. Paddle Email: [email protected] *Presented in part at the 26th Annual Congress of the European Society of Intensive Care Medicine, Paris, France; October 2013. Accepted: 26 April 2014
Introduction Acute kidney injury (AKI) requiring renal replacement therapy (RRT) is common in critically ill patients, with an incidence of up to 9.4% [1]. It is associated with a poor prognosis, with a hospital mortality of between 35% and 63.8% [2, 3]. Among survivors, there is a continued adverse impact on quality of life (QoL) [4]. Studies that have assessed the impact of AKI with or without RRT on QoL have used global measures that incorporate physical, pain, social and other parameters, as well as psychological morbidity [2, 5–12]. There is a paucity of data on the specific psychological impact of RRT. © 2014 The Association of Anaesthetists of Great Britain and Ireland
We sought to assess the impact of continuous RRT (CRRT) on the psychological health of survivors of critical illness. Our hypothesis was that the added burden of CRRT would increase the prevalence and severity of anxiety, depression and stress reactions in this group of patients, compared with a matched group of critical illness survivors who did not require CRRT.
Methods We conducted an observational study of survivors of critical illness who had received CRRT in our institution, and compared them with matched pairs of 1127
Anaesthesia 2014, 69, 1127–1132
survivors of critical illness who had not received CRRT. The study was approved by the Cornwall and Plymouth Research Ethics Committee, and performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki. As the study was purely observational, and used existing, routinely collected data, informed consent was not required from the subjects. We collected data on all patients receiving CRRT in our intensive care unit (ICU), which is a mixed adult medical and surgical general ICU, over a fiveyear period from 1 January 2007 to 31 December 2011. All patients receiving CRRT met the RIFLE (Risk, Injury, Failure, Loss and End-stage kidney disease) criteria for AKI. No patient was treated with intermittent haemodialysis. Patients who survived to three months after hospital discharge were sent a standardised QoL postal questionnaire incorporating the Hospital Anxiety and Depression Scale (HADS) and Impact of Events Scale (IES). We sought matched controls from critically ill patients on our unit who had not received CRRT. Each patient receiving CRRT was matched with a control patient according to the following criteria: admission to the critical care unit within one year of index case; same sex; age within five years; and APACHE II score within five points. This was defined as a good match. Where more than one match was identified, a random number generator (from within Microsoft Excel software) was used to select a single match. Where no match was found, the matching criteria were expanded to a control patient within 10 years of age, and within 10 APACHE II points. This was defined as a poor match. All critically ill patients in our institution are sent a postal questionnaire three months after hospital discharge, which incorporates the HADS and IES, and we used this to evaluate aspects of psychological morbidity. We used routinely collected data from our inpatient database for all demographic and survival data. We conducted multivariate analysis of matched pairs using a general linear model to compare HADS and IES between CRRT patients and controls. We used the Statistical Package for the Social Sciences software version 19 (SPSS Inc., Chicago, IL, USA) for this and for all descriptive statistics. Two-by-two chi-squared 1128
Paddle et al. | Psychological outcomes of dialysis in critical care
tests were used to compare numbers of patients with scores consistent with clinical anxiety, depression and impact of events. Statistical significance was assumed at p < 0.05.
Results During the study period, 395 patients received CRRT on our critical care unit, of which 342 met the RIFLE criteria for AKI. Of these, 179 (52.3%) survived to hospital discharge and 161 (47.1%) were alive at 90 days. All were sent questionnaires, and 77 (47.8% of survivors at 90 days) returned these completed. Figure 1 shows the CONSORT flow diagram for the study. Matches for these 77 patients were sought from our ICU database. Table 1 shows the baseline characteristics of the CRRT group and controls.
Enrollment
Assessed for eligibility (n = 395) Excluded • Not meeting RIFLE criteria for AKI (n = 53) Enrolled (n = 342) Died • Hospital death (n = 163) • Post-discharge, before 90 days (n = 18)
Survived (n = 161) Postal questionnaires sent to all No reply to questionnaire (n = 84)
Allocation
Database of ICU survivors with completed questionnaires (n = 338)
Analysis
Completed questionnaires (n = 77)
Matched pairs found for HADS analysis (n = 77)
Matched pairs found for IES analysis (n = 72)
Figure 1 CONSORT flow diagram. RIFLE, Risk, Injury, Failure, Loss and End-stage kidney disease; AKI, acute kidney injury; HADS, Hospital Anxiety and Depression Scale; IES, Impact of Events Scale. © 2014 The Association of Anaesthetists of Great Britain and Ireland
Paddle et al. | Psychological outcomes of dialysis in critical care
Table 1 Characteristics of patients who underwent continuous renal replacement therapy (CRRT) and control patients. Values are mean (95% CI) or number (proportion).
All matches n Age; years APACHE II score Male Good matches only n Age APACHE II score Male
CRRT
Control
77 63 (60–66) 24.3 (23.1–25.4) 45 (58.4%)
77 64 (61–67) 22.5 (21.4–23.6) 45 (58.4%)
65 66 (63–69) 24.2 (22.9–25.5) 40 (61.5%)
65 66 (63–69) 23.0 (21.8–24.1) 40 (61.5%)
p value
0.037 < 0.001
1.00
Anaesthesia 2014, 69, 1127–1132
Table 2 Hospital Anxiety and Depression Scale (HADS) and Impact of Events Scale (IES) in patients who underwent continuous renal replacement therapy (CRRT) and control patients. Values are mean (95% CI) or number (proportion).
HADS n Score Patients scoring ≥ 8 IES n Score Patients scoring ≥ 24
CRRT
Control
p value
77 11.4 [9.6–13.2] 49 [63.6%]
77 11.7 [9.7–13.7] 49 [63.6%]
0.846 1.00
72 20.1 [15.7–24.6] 24 [33.3%]
72 21.9 [16.2–26.2] 26 [36.1%]
0.752 0.86
< 0.001
The two groups were broadly comparable in terms of age and sex match. However, the APACHE II scores were significantly higher in the CRRT group. Although the difference was numerically smaller when only good matches were compared, a significant difference remained. This reflected the matching process that allowed for absolute differences in APACHE II score (as it would not have been possible to find matches with exactly the same scores in the database), and tended to select control patients that were slightly less sick. The multivariate analysis allowed for the assessment of the interaction of these variables within the HADS and IES scores where no significant interactions were found. Seventy-seven CRRT patients completed the questions required to calculate a HADS score, as did all 77 matched controls. Two of the CRRT patients did not complete sufficient questions to calculate an IES score, nor did three (different) control patients. The IES analysis was therefore conducted on 72 matched pairs. Table 2 illustrates the mean scores for the two tests, comparing CRRT patients with controls, and the number of patients with clinically relevant symptoms of anxiety, depression and post-traumatic stress reactions. We used a HADS score of 8 or more and an IES score of 24 or more as thresholds of clinical relevance [13, 14], and found no differences between the two groups for either test. © 2014 The Association of Anaesthetists of Great Britain and Ireland
Using a general linear model and multivariate analysis, accounting for any differences between the two groups with respect to age or APACHE II score, we found no significant differences between the HADS scores for CRRT patients and the matched controls. This remained the case whether we included all matches, or restricted the analysis to only the 65 good matches. Similarly, and using the same model, we found no significant differences between the IES scores in the two groups.
Discussion To the best of our knowledge, this is the first study that specifically assesses the impact of CRRT on the psychological health of critically ill patients in comparison with matched control patients who do not receive CRRT. Symptoms of anxiety and depression were common in our patients, affecting about two thirds of both patients receiving CRRT and matched controls. About a third of patients in both groups exhibited symptoms of clinically relevant post-traumatic stress. Similar high levels of psychiatric morbidity in critically ill patients have been found by other authors [15]. Among patients receiving haemodialysis for end-stage renal disease, symptoms of anxiety and depression are common, with 45.7% of subjects meeting criteria for an anxiety disorder [16] and 23.7% with significant depressive symptoms [17]. Symptoms consistent with post-traumatic stress are also common in this patient 1129
Anaesthesia 2014, 69, 1127–1132
group, with 77.8% of patients reporting at least one traumatic event [18]. Other studies have used global measures of QoL in this patient population, that incorporate domains pertinent to mental health. The Finnish Acute Kidney Injury study [11] assessed QoL in survivors of critical illness with AKI across ten centres in Finland, using the EuroQol-5D (EQ-5D) questionnaire. Subjects were assessed at six months. The EQ-5D assesses QoL across five health dimensions, of mobility, self-care, usual activities, pain or discomfort and anxiety or depression. The anxiety and depression domain asks only three questions and there is no assessment of potential impact of events on mental health. Another Finnish study [12] did look specifically at patients receiving RRT, but again used the EQ-5D assessment. Both studies found that QoL was poor by objective assessment in survivors of AKI, but that patients rated their subjective health to be as good as matched controls from the general population. Two European studies used the Short Form Health Survey (SF-36) questionnaire to report QoL in ICU patients with AKI [8, 9]. This assessment is also of global measures of health, covering eight health dimensions, of which mental health is just one, and the scores are aggregated. Again, no assessment is made of the potential impact of events. One study [8] reported a reduction in QoL at six months, the other [9], which repeated the assessment at different time intervals, reported a reduction in three domains of the SF-36 for ICU patients with AKI compared with those without at ICU discharge, but this difference was no longer present at six months. The Nottingham health profile, or simplified versions of it, has been used to assess QoL in three studies from Australia [5] and Europe [6, 7]. All were restricted to patients from ICUs receiving RRT. Two studies assessed QoL at six months; one study did not report the time interval. Again, this assessment measures a global QoL based on activity and functional health and is not specific to psychological health. All three studies reported that survivors had a good or acceptable QoL. We did not find any difference in measures of psychiatric morbidity between patients receiving CRRT and controls, and this remained the case for both 1130
Paddle et al. | Psychological outcomes of dialysis in critical care
simple chi-squared analysis and multivariate analysis that accounted for baseline differences. We therefore believe this study provides evidence that the added burden of RRT does not adversely affect the psychological health of critically ill patients. There may be a number of reasons for this. Firstly, AKI that is sufficiently severe to require RRT generally occurs during, and is part of, the critical illness, such that the patient is overwhelmed by stressors to the extent that RRT adds little to the perceived stress of the total illness period. Secondly, many of these patients will be sedated during their critical illness in order to facilitate treatment modalities such as mechanical ventilation, so the patient may not even be aware that he or she is receiving RRT. In our ICU, sedation is achieved with a combination of propofol and opioid infusion, titrated to a defined sedation target, and this applies both to patients who receive CRRT and to those who do not. Our findings are supported by the small number of other studies that have compared QoL outcomes in ICU patients with and without AKI [2, 9, 11]. None of these studies looked specifically at psychological morbidity, and only one [11] included a matched-pairs analysis. Even if RRT does increase psychological morbidity, it may be that this does not persist when the AKI resolves. While we did not specifically assess long-term dependence on RRT for this study, we know that in our own cohort of critically ill patients who receive RRT, only 3.7% still require RRT at 30 days [19], similar to an international finding of 5.3% of patients requiring RRT at hospital discharge [3]. Thus it may be that by three months after illness, any morbidity attributable to the period of RRT during critical illness has resolved for the overwhelming majority of patients who are no longer dialysis dependent. Conversely, it is possible that there is a difference in psychological morbidity between the two groups that our study failed to detect. Our study may have been under-powered, even though we sought to include all survivors over a five-year period. We chose assessment tools that have been validated in various patient groups [13, 14], including patient with endstage renal disease receiving haemodialysis or peritoneal dialysis [20] such that we would expect the HADS and IES scores to provide a better assessment of psychological trauma than a more generic score of QoL. © 2014 The Association of Anaesthetists of Great Britain and Ireland
Paddle et al. | Psychological outcomes of dialysis in critical care
The response rate to our questionnaires was relatively low, so there is a chance of selection bias. Patients suffering from psychiatric morbidity might be more inclined to respond as a means of seeking help. Alternatively, survivors with psychological sequelae might be less likely to respond, owing to the emotional distress that answering the questions might bring. We cannot exclude this potential bias. However, the response rate in this study is similar to the response rate we see for all ICU survivors (51.8%, unpublished clinic data), such that any bias would be expected to be matched in the control group. The timing of the survey may have influenced the responses, and some authors have observed that QoL improves over time following the illness [4]; however, that improvement may be restricted to physical and emotional domains of QoL, and not to mental health [4]. Other authors have failed to find any association between timing of survey and QoL [12]. In any event, both RRT and control groups were surveyed at the same three-month interval after discharge. Other limitations of our study should be noted. As it is a retrospective observational study, we are not able to comment on causality. As a single-centre study, our findings may not be applicable to other geographical areas or ethnic groups, however, our patients’ baseline demographics are similar to those of other critically ill patients in general hospitals across the UK. In conclusion, our five-year observational study of survivors of critical illness demonstrated that psychiatric morbidity, as assessed by HADS and IES, is common. However, our analysis found that in an ICU population matched for age, sex, and severity of illness, there was no additional psychological morbidity from RRT.
Acknowledgements The authors thank Jo Palmer from the Research and Development department of the Royal Cornwall Hospital for her assistance with statistical analysis.
Competing interests No external funding and no competing interests declared. © 2014 The Association of Anaesthetists of Great Britain and Ireland
Anaesthesia 2014, 69, 1127–1132
References 1. Nisula S, Kaukonen K-M, Vaara ST, et al. Incidence, risk factors and 90-day mortality of patients with acute kidney injury in Finnish intensive care units: the FINNAKI study. Intensive Care Medicine 2013; 39: 420–8. 2. Vaara ST, Pettil€a V, Reinikainen M, Kaukonen K-M. Populationbased incidence, mortality and quality of life in critically ill patients treated with renal replacement therapy: a nationwide retrospective cohort study in finnish intensive care units. Critical Care 2012; 16: R13. 3. Uchino S, Bellomo R, Morimatsu H, et al. Continuous renal replacement therapy: a worldwide practice survey. Intensive Care Medicine 2007; 33: 1563–70. 4. Dowdy DW, Eid MP, Sedrakyan A, et al. Quality of life in adult survivors of critical illness: a systematic review of the literature. Intensive Care Medicine 2005; 31: 611–20. 5. Gopal I, Bhonagiri S, Ronco C, Bellomo R. Out of hospital outcome and quality of life in survivors of combined acute multiple organ and renal failure treated with continuous venovenous hemofiltration/hemodiafiltration. Intensive Care Medicine 1997; 23: 766–72. 6. Korkeila M, Ruokonen E, Takala J. Costs of care, long-term prognosis and quality of life in patients requiring renal replacement therapy during intensive care. Intensive Care Medicine 2000; 26: 1824–31. 7. Morgera S, Kraft AK, Siebert G, Luft FC, Neumayer H-H. Longterm outcomes in acute renal failure patients treated with continuous renal replacement therapies. American Journal of Kidney Diseases 2002; 40: 275–9. 8. Delannoy B, Floccard B, Thiolliere F, et al. Six-month outcome in acute kidney injury requiring renal replacement therapy in the ICU: a multicentre prospective study. Intensive Care Medicine 2009; 35: 1907–15. 9. Hofhuis JGM, van Stel HF, Schrijvers AJP, Rommes JH, Spronk PE. The effect of acute kidney injury on long-term healthrelated quality of life: a prospective follow up study. Critical Care 2013; 17: R17. 10. Johansen KL, Smith MW, Unruh ML, Siroka AM, O’Connor TZ, Palevsky PM. Predictors of health utility among 60-day survivors of acute kidney injury in the veterans affairs/national institutes of health acute renal failure trial network study. Clinical Journal of the American Society of Nephrology 2010; 5: 1366–72. 11. Nisula S, Vaara ST, Kaukonen K-M, et al. Six-month survival and quality of life of intensive care patients with acute kidney injury. Critical Care 2013; 17: R250. €m A, Tallgren M, Peltonen S, R€as€anen P, Pettil€ 12. Ahlstro a V. Survival and quality of life of patients requiring acute renal replacement therapy. Intensive Care Medicine 2005; 31: 1222–8. 13. Bjelland I, Dahl AA, Haug TT, Neckelmann D. The validity of the hospital anxiety and depression scale. An updated literature review. Journal of Psychosomatic Research 2002; 52: 69–77. 14. Asukai N, Kato H, Kawamura N, et al. Reliability and validity of the Japanese-language version of the impact of event scalerevised (IES-R-J): four studies of different traumatic events. Journal of Nervous and Mental Disease 2002; 190: 175–82. 15. Jackson JC, Hart RP, Gordon SM, Hopkins RO, Girard TD, Ely WE. Post-traumatic stress disorder and post-traumatic stress symptoms following critical illness in medical intensive care unit patients: assessing the magnitude of the problem. Critical Care 2007; 11: R27.
1131
Anaesthesia 2014, 69, 1127–1132 16. Cukor D, Coplan J, Brown C, et al. Anxiety disorders in adults treated by hemodialysis: a single-center study. American Journal of Kidney Diseases 2008; 52: 128–36. 17. Agganis BT, Weiner DE, Giang LM, et al. Depression and cognitive function in maintenance hemodialysis patients. American Journal of Kidney Diseases 2010; 56: 704–12. 18. Tagay S, Kribben A, Hohenstein A, Mewes R, Senf W. Posttraumatic stress disorder in hemodialysis patients. American Journal of Kidney Diseases 2007; 50: 594–601.
1132
Paddle et al. | Psychological outcomes of dialysis in critical care 19. Hayes-Bradley C, Caddel S, Paddle J. Observational cohort of renal replacement therapy patients at a district general hospital ICU: case mix and outcome. Critical Care 2009; 13: P272. 20. Loosman WL, Siegert CE, Korzec A, Honig A. Validity of the hospital anxiety and depression scale and the Beck depression inventory for use in end-stage renal disease patients. British Journal of Clinical Psychology 2010; 49: 507–16.
© 2014 The Association of Anaesthetists of Great Britain and Ireland
