Accurate Identification and Prognostication of End-of-Life State in Japan Using A Guideline-Based Diagnostic Method: A Diagnostic and Prognostic Study

doi:10.21203/rs.3.rs-1352779/v1

Background: Prognostic uncertainty can be a barrier to providing palliative care. Accurate prognostic estimation in patients at the end of life is a challenging issue.

Methods: A retrospective longitudinal observational study was conducted collaborated by three medical facilities in a rural super-aged community in Japan. We established a unique end-of-life diagnostic process consisting of (1) physicians’ judgement, (2) disclosure to patients, and (3) discussions at an end-of-life case conference (EOL-CC), which were based on Japanese end-of-life related guidelines. Participants were consecutive patients discussed at EOL-CC between January 1, 2010, and September 30, 2017. The primary study outcome was mortality within 6 months after the initial EOL-CC decision. The secondary outcome was diagnostic odds ratio using EOL-CC diagnosis (end-of-life or non-end-of-life) as an index test and overall survival (less than 6 months or 6 months and more) as a reference standard.

Results: A total of 315 patients were eligible for survival analysis (189 women, median [range] age 89 [54–107] years). In the survival analysis, end-of-life diagnosis was associated with much lower survival rate at 6 months after EOL-CC than non-end-of-life diagnosis (6.9% vs 43.5%, respectively; P < 0.001). However, 10 of the end-of-life patients were alive more than 1 year later (maximum 6.3 years). Of participants, 297 were eligible for diagnostic accuracy analysis (165 women, age 89 [54–107] years). The EOL-CC diagnosis showed high sensitivity (0.95; 95% CI 0.92–0.97) but low specificity (0.35; 95% CI 0.20–0.53) against the outcomes. It also showed the highest negative predictive value (0.44; 95% CI 0.23–0.64), the lowest negative likelihood ratio (0.14; 95% CI 0.07–0.29), and the highest diagnostic odds ratio (10.32; 95% CI 4.08–26.13) compared with other prognostic prediction tools (Charlson Comorbidity Index, Palliative Performance Scale, and Advanced Dementia Prognostic Tool).

Conclusions: The diagnostic process using the Japanese end-of-life guidelines enabled more accurate identification and prognostication of end of life than existing prognostic tools. However, the study population was limited to patients with severe conditions, thus further research is necessary to build full consensus.

Diagnostic accuracy

End of life

Prognostication

Overall survival

Diagnostic odds ratio.

Accurately estimating life expectancy in patients at the end of life (EOL) is extremely difficult, especially for patients with non-cancer diseases. One longitudinal study indicated that the progression of disability in the last year of life does not follow a predictable pattern [1]. In the last decade, several tools have been developed to predict prognosis of patients near the EOL; however, these tools lack accuracy [2]. Therefore, many older people close to death cannot be given an accurate prognosis [3]. Moreover, prognostic uncertainty can be a barrier to providing palliative care [4].

In Japan, EOL-associated issues have attracted nationwide attention because of several well-publicized cases of ventilator removal in 2006 [5]. In 2007, the Japanese Ministry of Health, Labour and Welfare (JMHLW) developed guidelines to establish the process of determining EOL status; these guidelines were revised in 2018 [6, 7]. The JMHLW guidelines described three main scenarios of the decision-making procedure: (1) When the patient’s intention can be confirmed, the patient must be involved in decision making about the treatment plan, following sufficient consensus-building discussions between the patient and multidisciplinary medical team. (2) When the patient’s intention cannot be confirmed, it should be gauged and respected, and the best possible plan put in place for the patient. (3) When sufficient consent cannot be obtained, the patient’s conditions and treatment plan must be discussed by a multidisciplinary team that includes experts from the medical field and other fields (Additional file 1: Text S1) [7]. This latter scenario emphasizes that patients should not be diagnosed as EOL by a single doctor; rather, several doctors should be involved, and the diagnosis should always be supported by professionals from a range of fields. Several subsequent EOL guidelines have been published in Japan [8, 9]. However, these guidelines do not provide any eligible criteria for end-stage diseases, unlike the National Hospice Organization Medical Guidelines [10], and their usefulness has not been validated. The difficulty in EOL diagnosis in non-cancer diseases may arise from prognostic uncertainty, as shown by several Japanese studies [11]. This can lead to inappropriate therapy and use of invasive techniques such as tube feeding without adequate evaluation of patients’ conditions [12].

However, in the authors’ institutions, a rigorous diagnostic process based on the Japanese EOL-related guidelines has been used for over a decade to diagnose many older or critically ill patients as EOL. This process and the associated EOL diagnostic criteria build on the existing guidelines by including a requirement to confirm the irreversibility of the patient’s condition. We conducted this study to retrospectively review cases diagnosed as EOL or non-EOL and to evaluate their diagnostic accuracy. The hypothesis was that EOL diagnosis using our rigorous diagnostic process, which follows the Japanese guidelines, correctly predicts patients’ prognosis.

This study was approved by the institutional review board of Nanto Municipal Hospital (No. Shiminbyouin-96). In accordance with the guidelines of the Japanese Ministry of Education, Culture, Sports, Science and Technology [13], the requirement for obtaining participant informed consent was waived. This study involved no intervention for participants and followed the Standards for Reporting of Diagnostic Accuracy reporting guideline [14].

Study Setting

The study was a collaboration between several medical facilities in Nanto, a super-aged city. In 2017, 36.7% of the population of Nanto were aged 65 years or above. To respond appropriately to the large number of deaths in this super-aged population, we required a rigorous method of making EOL judgments to shift the emphasis of medical care from acute care to terminal care. Using Japanese EOL-related guidelines, we established a unique diagnostic method in 2007 (Fig. 1). The diagnostic method included the following three processes: (1) Attending physician judges that the patient is in the EOL state. (2) Attending medical team disclose the patient’s medical condition and present treatment options, including life-prolonging methods, to the patient and his or her family. (3) The patient’s condition and treatment plan must be discussed in an EOL case conference (EOL-CC) between experts from different medical fields. After there, a final diagnosis is made in the EOL-CC (Fig. 1). Most patients discussed at the EOL-CC were being followed up at one of our institutions (Nanto Municipal Hospital, Nanto Family and Community Medical Center, or Toga Clinic).

Study Design and Participants

This was a retrospective longitudinal observational study conducted to assess the diagnostic accuracy in determining whether a critically ill patient is in an EOL state. Participants were consecutive patients presented at an EOL-CC, and for whom a decision had been made, between January 1, 2010, and September 30, 2017. Each patient’s attending physician determined the necessity for discussion of their case at the EOL-CC. Participants were retrospectively extracted by referring to a database that recorded details of the EOL-CC. This database contained several types of information, including patient ID, age, sex, clinical diagnosis, EOL-CC discussions, and EOL-CC decisions. If a patient had never received an EOL diagnosis (non-EOL), the reason was noted. Participants were excluded from the diagnostic accuracy analysis if an EOL-CC decision was pending. For cases that had been discussed repeatedly, only the initial EOL-CC decision was included.

Diagnostic Criteria for EOL State

At every EOL-CC, all the following conditions had to be fulfilled to diagnose a target patient as EOL: (1) the patient had received the best medical care for recovery and no curable interventions remained; (2) even if treatment was continued, the patient’s prognosis was poor (irreversible condition) and death was likely to be imminent; the estimated survival was less than 6 months; (3) several doctors, nurses, co-medical staffs, the patient, and his or her family recognized (1) and (2), which were rigorously confirmed according to the medical findings and in-depth discussions (Fig. 1). In particular, specialists from different fields provided prognoses for patients with malignancy (Additional file 1: Table S1). These EOL criteria follow the Japanese EOL-related guidelines [6–9]. Type of trajectory to EOL (EOL-type) was determined according to presentations by the attending physicians and medical records, and was categorized as one of the following: terminal illness (patients with advanced cancer), organ failure, frailty (patients with advanced dementia), and unclassifiable (unable to be classified into the first three categories), with reference to a previous study [15].

Data collection and follow-up

Data on baseline characteristics (symptoms, histories, comorbidities, laboratory data, medications, physical and cognitive functions, and others) were obtained from the EOL-CC database and medical records. Retrospective scores on the Charlson Comorbidity Index (CCI), the Palliative Performance Scale (PPS), and the Advanced Dementia Prognostic Tool (ADEPT) were analyzed in this study [16–18]. The clinical courses and outcomes for all participants were retrospectively reviewed using their medical records.

Outcomes

The primary study outcome was mortality within 6 months and the secondary outcome was diagnostic odds ratio (DOR). Overall survival (OS) was defined as days between the initial EOL-CC decision (EOL, non-EOL, or pending) and death from any cause. True-EOL was defined as OS less than 6 months, whereas OS 6 months or more was defined as other-than-EOL. If a patient had transferred to another institution outside our local medical area, the follow-up was censored at the day of the last medical record made when the patient was alive. If a follow-up period was less than 6 months during survival, the participant was excluded from analysis and recorded as having missing data.

Statistical analysis

To evaluate the diagnostic accuracy of our method of determining EOL, we used OS as the reference standard. To obtain an index value of diagnostic accuracy of the EOL-CC decision against the primary outcome (true-EOL or other-than-EOL), we calculated sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), positive likelihood ratio (PLR), negative likelihood ratio (NLR), and DOR [19]. OS was analyzed by survival analysis using the Kaplan–Meier method and was statistically compared using the log-rank test. Logistic regression was used to evaluate the effect of potential confounding factors on death within 6 months. Factors were selected for multivariate analysis if the P-value in the univariate analysis was < 0.20.

All numerical variables were tested for distribution normality using the Kolmogorov–Smirnov test, and data with two-sided P-values of < 0.05 were regarded as non-normally distributed. For two-group comparisons, normally distributed variables and non-normally distributed variables were compared using the t-test and Mann–Whitney U test, respectively. For comparisons among more than two groups, post hoc multiple comparisons were made using Tukey’s honestly significant difference test after confirming heterogeneity of means using one-way analysis of variance. For categorical data, the proportions in each group were compared using Fisher’s exact test. These statistical analyses were performed using XLSTAT version 2021.4.1 (Addinsoft Inc. Paris, France, https://www.xlstat.com/ja/) and EZR version 1.54 (Saitama Medical Center, Jichi Medical University, Saitama, Japan). EZR is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria, version 4.0.3) [20]; it is a modified version of R commander (version 2.7-1) designed to provide additional statistical functions frequently used in biostatistics. A two-sided P-value < 0.05 was considered statistically significant.

Study population and baseline characteristics

From January 1, 2010, to September 30, 2017, a total of 330 cases were discussed at the EOL-CC. We identified 319 participants who had received a first decision by the EOL-CC. The EOL-CC evaluated 275 (86%) patients as EOL, 24 (8%) as non-EOL, and 20 (6%) as pending. Of these 319 participants, 315 were eligible for the retrospective analysis (Fig. 2). The baseline characteristics and EOL types of the study population are shown in Table 1. Median (range) age of all participants was 89 (54–107) years. The terminal illness group was significantly younger than the other groups. The frailty group was characterized by severe deterioration in functions such as physical, cognitive, and eating abilities. Most patients (85%) had eating problems, mainly owing to reduced oral intake.

Table 1

Baseline characteristics
Variables	All (n = 315)	Terminal illness (advanced cancer) (n = 139)	Frailty (advanced dementia) (n = 106)	Organ failure (n = 41)	Unclassifiable cases (n = 29)
Age, median (range), y	89 (54–107)	85 (54–101)	91 (70–107)	91 (70–107)	92 (79–99)
Women, No. (%)	189 (60)	73 (53)	74 (70)	24 (59)	18 (62)
Hospitalized patients, No. (%)	246 (78)	102 (73)	84 (79)	35 (85)	25 (86)
CCI, median (range)	8 (4–17)	10 (5–17)	6 (4–11)	7 (5–13)	7 (5–11)
Physical functions
Barthel Index, No. (%)
0	184 (58)	44 (32)	96 (91)	21 (51)	23 (79)
5–80	94 (30)	63 (45)	10 (9)	15 (37)	6 (21)
85–100	8 (3)	7 (5)	0	1 (2)	0
NA	29 (9)	25 (18)	0	4 (10)	0
^aBedridden level, No. (%)
Level J or A	16 (5)	13 (9)	0	2 (5)	1 (3)
Level B or C	299 (95)	126 (93)	106 (100)	39 (95)	28 (97)
Cognitive functions
FAST, No. (%)
≤6	84 (27)	73 (52)	0	11 (27)	0
6a–6e	30 (10)	17 (12)	2 (2)	3 (7)	8 (28)
7a–7c	46 (15)	10 (7)	24 (23)	8 (20)	4 (14)
7d–7f	117 (37)	14 (10)	77 (73)	12 (29)	14 (48)
NA	38 (12)	25 (18)	3 (3)	7 (17)	3 (10)
HDS-R, No. (%)
0	200 (63)	51 (37)	103 (97)	22 (54)	24 (83)
1–19	42 (13)	27 (20)	3 (3)	8 (20)	4 (14)
20–30	9 (3)	6 (4)	0	2 (5)	1 (3)
NA	64 (20)	55 (40)	0	9 (22)	0
^aRating of Dementia, No. (%)
Normal	23 (7)	22 (16)	0	1 (2)	0
Level I or II	88 (28)	65 (47)	4 (4)	14 (34)	5 (17)
Level III, IV or M	204 (65)	52 (37)	102 (96)	26 (63)	24 (83)
Eating and swallowing functions
^bEating problems, No. (%)	268 (85)	100 (72)	105 (99)	35 (85)	28 (97)
Oral intake, No. (%)
0	94 (30)	22 (16)	38 (36)	17 (41)	17 (59)
1–500 kcal/day	116 (37)	46 (33)	52 (49)	10 (24)	8 (28)
>500 kcal/day	68 (22)	46 (33)	11 (10)	9 (22)	2 (7)
Dependence on PEG or NG	4 (1)	1 (1)	1 (1)	1 (2)	1 (3)
NA	33 (10)	24 (17)	4 (4)	4 (10)	1 (3)
PPS, No. (%)
10	106 (34)	25 (18)	43 (41)	19 (46)	19 (66)
20	105 (33)	49 (35)	40 (38)	10 (24)	6 (21)
30	46 (15)	18 (13)	23 (22)	3 (7)	2 (7)
40	47 (15)	38 (27)	0	7 (17)	2 (7)
≥50	11 (3)	9 (6)	0	2 (5)	0
ADEPT, mean (SD)	16.6 (3.6)	15.4 (3.5)	17.5 (2.9)	17.7 (4.3)	17.4 (3.5)
Residence before EOL-CC, No. (%)
Home	189 (60)	109 (78)	28 (26)	30 (73)	22 (76)
Nursing facilities	117 (37)	25 (18)	75 (71)	10 (24)	7 (24)
Long-term hospitals	8 (3)	5 (4)	3 (3)	0	0
Other hospitals	1 (0)	0	0	1 (2)	0
Footnote
^aThe scoring methods “Bedridden level” and “Rating of Dementia” are defined by the Japanese Ministry of Health, Labour and Welfare; details are available at http://www.mhlw.go.jp/english/database/db-hss/dl/siel-2010-04.pdf.
^bEating problems were defined as swallowing or chewing problems (need to change food types), behavior problems such as refusal to eat or drink, dependence on help for eating, suspected dehydration, and persistently reduced oral intake.
Abbreviations: ADEPT, Advanced Dementia Prognostic Tool; CC, case conference; CCI, Charlson Comorbidity Index; EOL, end of life; FAST, Functional Assessment Staging Test; HDS-R, Hasegawa Dementia Scale – Revised; kcal, kilocalorie; NA, not available; NG, nasogastric tube; PEG, percutaneous endoscopic gastrostomy; PPS, Palliative Performance Scale; SD, standard deviation.

Associations between EOL-CC diagnosis and primary outcome

Follow-up data were collected between October 1, 2018, and December 2019. All participants were already dead before this retrospective study began. The results of the survival analysis using the Kaplan–Meier method are shown in Fig. 3. Non-EOL diagnosis compared with EOL diagnosis was associated with a higher survival rate at 6 months after EOL-CC (43.5% vs. 6.9%; log-rank P < 0.001). However, a small number of EOL patients were alive more than 1 year later, (maximum of 6.3 years) (Fig. 3A). In EOL patients, EOL type was associated with a significant difference in survival at 6 months after the EOL-CC (terminal illness 2.2%, 95% CI 0.65–5.8; frailty 16.1%, 95% CI 9.3–24.5; organ failure 2.9%, 95% CI 0.2–12.7; unclassifiable cases 6.7%, 95% CI 0.4–26.0; log-rank P = 0.019, see Fig. 3B).

Diagnostic odds ratio and other index values of diagnostic accuracy

Of participants, 297 were eligible for the diagnostic accuracy analysis after excluding patients pending a decision (women 179, median [range] age 89 [54–107] years). Participant baseline characteristics were similar to those of the survival analysis population (Additional file 1: Table S2). The cross-tabulation of EOL-CC decision (EOL or non-EOL) by the reference standard results (true-EOL or other-than-EOL) is shown in Table 2. The index test and reference standard results by EOL type distribution are shown in Table 3. The frailty group had the lowest proportion of EOL diagnoses (87/98; 89%) and the highest false positive rate (other-than-EOL in the EOL patients, 14/87, 16%) of all EOL types. Estimates of diagnostic accuracy and their precision for the EOL-CC diagnosis are shown in Table 4. For all participants, the EOL-CC EOL diagnosis had high sensitivity (0.95, 95% CI 0.92–0.97) and PPV (0.93, 95% CI 0.90–0.96) but low specificity (0.35, 95% CI 0.20–0.53). It also had the highest NPV (0.44, 95% CI 0.23–0.64), the lowest NLR (0.14, 95% CI 0.07–0.29), and substantially higher DOR (10.32, 95% CI 4.08–26.13) compared with the other prognostic prediction tools, even though any cutoff values were used for these. In the subgroup analysis by EOL type, the PPV and NPV of the terminal illness group and organ failure group were extremely high, but those of the frailty group were relatively low (Table 4).

Table 2

Cross-tabulation of index test results by reference standard results
Indicator	Indicator positive/OS < 6M (true positive)	Indicator positive/OS ≥ 6M (false positive)	Indicator negative/OS < 6M (false negative)	Indicator negative/OS ≥ 6M (true negative)
Diagnosis by EOL-CC	255	19	13	10
Terminal illness	134	3	0	1
Frailty	73	14	7	4
Organ failure	34	1	0	2
CCI
≥6	256	25	26	8
≥7	221	16	61	17
≥8	178	9	104	24
≥9	138	5	144	28
≥10	107	5	175	28
PPS
≤10	99	7	183	26
≤20	193	18	89	15
≤30	229	28	53	5
≤40	272	32	10	1
ADEPT
≥15	185	25	90	8
≥16	157	16	118	17
≥17	130	11	145	22
≥18	111	7	164	26

Abbreviations: 6M, 6 months; ADEPT, Advanced Dementia Prognostic Tool; CC, case conference; EOL, end of life; OS, overall survival; PPS, Palliative Performance Scale.

Table 3

Summary of index test results and reference standard results by EOL type
Index test results (EOL-CC diagnosis), standard reference (the primary outcome)	Terminal illness (n = 138)	Frailty (n = 98)	Organ failure (n = 37)	Unclassifiable cases (n = 24)	P
EOL (diagnosed by EOL-CC)	137	87	35	15	< 0.001
Non-EOL (diagnosed by EOL-CC)	1	11	2	9	< 0.001
True positive; True-EOL (OS < 6M) in EOL	134	73	34	14	< 0.001
False positive; Other-than-EOL (OS ≥ 6M) in EOL	3	14	1	1	< 0.001
False negative; True-EOL (OS < 6M) in non-EOL	0	7	0	6	0.26
True negative; Other-than-EOL (OS ≥ 6M) in non-EOL	1	4	2	3	0.26

Abbreviations:6M, 6 months; CC, case conference; EOL, end of life; OS, overall survival.

Table 4

Statistics for diagnostic accuracy
Index values	Sensitivity (95% CI)	Specificity (95% CI)	PPV (95% CI)	NPV (95% CI)	PLR (95% CI)	NLR (95% CI)	DOR (95% CI)
EOL diagnosis by EOL-CC ―All patients	0.95 (0.92–0.97)	0.35 (0.20–0.53)	0.93 (0.90–0.96)	0.44 (0.23–0.64)	1.45 (1.11–1.89)	0.14 (0.07–0.29)	10.32 (4.08–26.13)
EOL diagnosis by EOL-CC ― Terminal illness group	1.00 (0.97–1.00)	0.25 (0.04–0.71)	0.98 (0.95–1.00)	1.00 (1.00–1.00)	1.33 (0.76–2.35)	0.00	Inf
EOL diagnosis by EOL-CC ― Frailty group	0.91 (0.83–0.96)	0.22 (0.09–0.46)	0.84 (0.76–0.92)	0.36 (0.08–0.65)	1.13 (0.91–1.52)	0.39 (0.13–1.20)	2.90 (0.81–10.90)
EOL diagnosis by EOL-CC ― Organ failure group	1.00 (0.88–1.00)	0.67 (0.21–0.94)	0.97 (0.92–1.00)	1.00 (1.00–1.00)	3.00 (0.61–14.86)	0.00	Inf
CCI ≥ 6	0.91 (0.87–0.94)	0.28 (0.15–0.46)	0.92 (0.88–0.94)	0.26 (0.10–0.41)	1.26 (1.00–1.59)	0.31 (0.15–0.63)	4.06 (1.65–9.98)
CCI ≥ 7	0.80 (0.75–0.84)	0.45 (0.29–0.62)	0.93 (0.90–0.96)	0.19 (0.10–0.29)	1.45 (1.04–2.02)	0.45 (0.28–0.72)	3.22 (1.48–7.01)
CCI ≥ 8	0.65 (0.59–0.70)	0.69 (0.51–0.83)	0.95 (0.92–0.98)	0.17 (0.11–0.24)	2.08 (1.20–3.60)	0.51 (0.38–0.69)	4.05 (1.80–9.09)
CCI ≥ 9	0.50 (0.44–0.56)	0.83 (0.65–0.93)	0.96 (0.93–1.00)	0.15 (0.10–0.21)	2.92 (1.31–6.54)	0.60 (0.49–0.74)	4.87 (1.88–12.66)
CCI ≥ 10	0.39 (0.33–0.45)	0.83 (0.65–0.93)	0.95 (0.92–0.99)	0.13 (0.08–0.18)	2.25 (1.00–5.07)	0.74 (0.61–0.90)	3.04 (1.17–7.92)
PPS = 10	0.35 (0.29–0.41)	0.83 (0.65–0.93)	0.95 (0.91–0.99)	0.12 (0.08–0.17)	2.01 (0.89–4.54)	0.79 (0.42–0.97)	2.08 (0.97–4.45)
PPS ≤ 20	0.69 (0.63–0.74)	0.48 (0.31–0.66)	0.93 (0.89–0.96)	0.14 (0.07–0.21)	1.34 (0.93–1.91)	0.64 (0.64–1.05)	1.81 (0.88–3.71)
PPS ≤ 30	0.81 (0.76–0.85)	0.17 (0.07–0.35)	0.90 (0.86–0.94)	0.09 (0.02–0.16)	0.98 (0.82–1.17)	1.10 (0.48–2.54)	0.89 (0.34–2.35)
PPS ≤ 40	0.97 (0.94–0.98)	0.03 (0.00–0.19)	0.90 (0.87–0.94)	0.10 (0.00–0.29)	1.00 (0.93–1.08)	0.97 (0.13–7.42)	1.03 (0.18–6.00)
ADEPT ≥ 15	0.67 (0.61–0.73)	0.24 (0.12–0.43)	0.89 (0.84–0.93)	0.08 (0.02–0.13)	0.88 (0.71–1.10)	1.37 (0.70–2.66)	0.65 (0.27–1.54)
ADEPT ≥ 16	0.57 (0.51–0.63)	0.55 (0.38–0.72)	0.92 (0.88–0.96)	0.13 (0.07–0.18)	1.27 (0.84–1.93)	0.78 (0.54–1.11)	1.64 (0.77–3.50)
ADEPT ≥ 17	0.48 (0.42–0.54)	0.69 (0.51–0.83)	0.93 (0.89–0.98)	0.13 (0.08–0.18)	1.53 (0.88–2.67)	0.76 (0.58–1.00)	2.01 (0.90–4.50)
ADEPT ≥ 18	0.40 (0.35–0.46)	0.79 (0.61–0.90)	0.95 (0.90–0.99)	0.13 (0.08–0.18)	1.94 (0.94–4.03)	0.75 (0.61–0.93)	2.58 (1.05–6.37)
Footnote
Each statistic was calculated using the findings in the leftmost column as a positive index test result and the primary outcome as the reference standard. For example, if a patient with an EOL diagnosis had the outcome of true-EOL or other-than-EOL, the patient was regarded as a true-positive or false-positive case, respectively. Patients determined as “pending” at the EOL-CC were excluded from this analysis.

Abbreviations: ADEPT, Advanced Dementia Prognostic Tool; ADL, activities of daily living; BI, Barthel Index; CC, case conference; CCI, Charlson Comorbidity Index; CI, confidence interval; DOR, diagnostic odds ratio; EOL, end of life; FAST, Functional Assessment Staging Test; HDS-R, Hasegawa Dementia Scale – Revised; Inf, positive infinity; N, number; NLR, negative likelihood ratio; NPV, negative predictive value; PLR, positive likelihood ratio; PPV, positive predictive value; PPS, Palliative Performance Scale.

The results of the logistic regression analysis of predictors of OS more than 6 months are shown in Table 5. The factors included a combination of performance on the functional measurement tools to indicate elements of condition severity. Age, CCI, terminal illness, and inability to communicate were found to be significant prognostic factors in the univariate analysis. However, multivariate analysis showed that only severe eating problem (oral intake < 500 kcal/day) was a significant prognostic factor (OR 0.27, 95% CI 1.10–0.74; P = 0.01).

Table 5

Logistic regression analysis of overall survival
Variables	Univariate OR (95% CI)	P	Multivariate OR (95% CI)	P
Age	1.06 (1.01–1.11)	0.03	1.02 (0.95–1.08)	0.61
Sex, male	0.53 (0.24–1.18)	0.12	0.50 (0.27–1.93)	0.51
CCI	0.72 (0.60–0.87)	< 0.001	0.79 (0.60–1.05)	0.11
Barthel Index (BI)	0.99 (0.97–1.01)	0.22	Not applicable	-
Bedridden Level, C	0.52 (0.22–1.18)	0.12	Not applicable	-
FAST, >7c	1.77 (0.84–3.76)	0.14	Not applicable	-
HDS-R	0.99 (0.93–1.06)	0.80	Not applicable	-
Rating of Dementia, III, IV, or M	3.37 (1.26–9.00)	0.02	Not applicable	-
PPS	1.02 (0.99–1.05)	0.27	Not applicable	-
ADEPT	0.98 (0.89–1.09)	0.72	Not applicable	-
Type of trajectory to EOL
Terminal illness (advanced cancer)	0.19 (0.04–0.79)	0.02	0.22 (0.03–1.60)	0.14
Frailty (advanced dementia)	1.54 (0.49–4.92)	0.46	1.10 (0.31–3.86)	0.88
Organ failure	0.68 (0.15–2.96)	0.60	0.31 (0.05–2.03)	0.22
Severe condition, element 1 (Almost bedridden); BI = 0, or Bedridden Level = C.	1.35 (0.56–3.24)	0.50	0.41 (0.09–1.80)	0.24
Severe condition, element 2 (Unable to communicate); HDS-R = 0, Rating of Dementia = IV or M, or FAST > 7c.	4.06 (1.39–11.9)	0.01	2.16 (0.34–13.8)	0.42
Severe condition, element 3 (Severe eating problem); oral intake < 500 kcal/day	0.58 (0.26–1.33)	0.20	0.27 (0.10–0.74)	0.01
Severe conditions, all elements 1–3	1.24 (0.58–2.67)	0.58	Not applicable	-

Abbreviations: ADEPT, Advanced Dementia Prognostic Tool; BI, Barthel Index; CCI, Charlson Comorbidity Index; CI, confidence interval; EOL, end of life; FAST, Functional Assessment Staging Test; HDS-R, Hasegawa Dementia Scale – Revised; OR, odds ratio; PPS, Palliative Performance Scale.

Previous studies indicate the lack of adequate scoring systems and diagnostic criteria to accurately determine EOL [2, 3]. Additionally, the Japanese EOL-related guidelines do not provide specific EOL diagnostic criteria, but recommend expert discussion of EOL based on findings such as severe functional decline, irreversibility of condition, and estimated prognosis [6–9]. Therefore, we established a process to diagnose EOL following these guidelines. Our EOL diagnosis showed high sensitivity, low NLR, and the highest DOR, but low specificity.

The patients presented at the EOL-CC were bedridden (unable to move independently), had severe dementia (unable to communicate), and reduced oral intake (unable to eat sufficiently), especially those in the frailty group (Table 1). Therefore, these three elements may prompt the attending physician to diagnose EOL. However, fulfillment of these three elements is not a prognostic factor (Table 5). This may be because the EOL diagnosis by the EOL-CC depended not only on these three elements but also on the reversibility of the patient’s condition (Fig. 1). Hui et al identified a key defining feature of EOL as life-limiting disease with irreversible decline [21]. However, such irreversibility has not been considered and incorporated into previous research [2, 16–18]. The novelty of our study is the use of the criterion of irreversibility, which may have improved diagnostic accuracy. Several Japanese EOL-related guidelines describe the need to take into account irreversibility before diagnosing EOL [8, 9, 22]. However, the criteria for irreversibility remain unclear and require clarification. Severe eating problem (oral intake < 500 kcal/day), one of the three elements of EOL diagnosis, had the greatest effect on participants’ prognosis (Table 5). Similar findings have been reported by previous studies [2, 3, 23, 24]. Accurate evaluation and appropriate intervention for eating problems is a challenging issue in this field [25]; however, addressing this issue would facilitate more accurate evaluation of irreversibility.

The PPV and NPV of the frailty group were relatively lower than those of the terminal illness and organ failure groups for the EOL-CC diagnosis (Table 4). In addition, the survival rate at 6 months after EOL-CC in the frailty group was significantly higher than that in the terminal illness group (16.1% vs. 2.2%, respectively, log-rank P < 0.001, see Fig. 3B). These findings indicate that accurate prognosis prediction of non-cancer patients with severe conditions is difficult. Recent studies have shown similar findings; some improvement of prognostication has been achieved in cancer fields [26, 27], but not in non-cancer fields [2, 3, 28, 29].

A recent systematic review suggested that ADEPT is a reliable prognostic tool for predicting risk of death at 6 months among persons with advanced dementia in nursing homes [30]. However, our EOL diagnostic method had a higher DOR (10.32, 95% CI 4.08–26.13) than ADEPT score ≥ 18 (2.58, 95% CI 1.05–6.37) in our study populations (Table 4). However, a slight risk of EOL misdiagnosis remained, especially in non-cancer patients; approximately 5% of EOL patients diagnosed by EOL-CC were alive more than 1 year later (Fig. 3A). In addition, the specificity was low because of the rigorous criteria for EOL diagnosis. Therefore, only some patients would benefit from the EOL diagnostic method presented here. To improve the usefulness of this diagnostic method, further research is needed to revise and validate it for new populations.

Study limitations

This study had some limitations. First, the cause of death was unknown for many participants, because the death certificate information could not be confirmed. It is possible that some patients died from causes different to the pathological conditions discussed at the EOL-CC. Future research should address this point. Second, some baseline data were missing because this was a retrospective study. The CCI, PPS, and ADEPT were scored retrospectively; therefore, their scores may not have been completely accurate. Third, individual differences in interventions such as nursing care, social resources, and medical treatments may have affected prognosis, but these were not investigated in this study. Patients with an estimated poor prognosis are less likely to receive burdensome interventions [31]. Similarly, some patients diagnosed as EOL may not be given the opportunity to recover from severe conditions. To prevent these problems, we designed a rigorous process by which to determine EOL based on several EOL-related guidelines. Therefore, the risk of misdiagnosis was minimal. Fourth, the high probability and high DOR results may reflect the attending physicians’ selection of patients to present at the EOL-CC. It is possible that physicians avoided presenting some patients owing to advance directives or advance care planning (ACP). It may have been assumed that EOL-CC decisions interfere with treatment strategies already determined by ACP, even though the usefulness of ACP at the EOL has not been proven [32, 33]. Therefore, these findings are only applicable to a limited range of clinical settings.

The diagnostic process we developed, following the Japanese EOL-related guidelines, was associated with more accurate identification and prognostication of EOL than existing prognostic tools. However, the study population was limited to patients with severe conditions. Furthermore, a slight risk of being misdiagnosed as EOL remained for non-cancer patients. Therefore, it is necessary to collect more data from clinical practice and attempt to improve the accuracy of diagnosis.

ACP

Advance care planning

ADEPT

Advanced Dementia Prognostic Tool

CCI

Charlson Comorbidity Index

DOR

Diagnostic odds ratio

End of life

EOL

EOL-CC

End-of-life Case Conference

JMHLW

Japanese Ministry of Health, Labour and Welfare

NLR

Negative likelihood ratio

NPV

Negative predictive value

OS

Overall survival

PLR

Positive likelihood ratio

PPS

Palliative Performance Scale

PPV

positive predictive value (PPV).

Ethics approval and consent to participate

This study was approved by the institutional review board of Nanto Municipal Hospital (No. Shiminbyouin-96). In accordance with the guidelines of the Japanese Ministry of Education, Culture, Sports, Science and Technology, the requirement for obtaining participant informed consent was waived.

Consent for publication

Not applicable

Availability of data and material

The datasets used and/or analyzed during this study are available from the corresponding author on reasonable request.

Competing interests

All authors declare that they have no competing interests.

Funding

None

Authors’ contributions

MA drafted the manuscript. MA, HA, and EM collected, analyzed, and interpreted the data. HA and EM substantively revised the manuscript. SM and YS contributed to study design and assisted in preparing the manuscript. All authors read and approved the final manuscript.

Acknowledgments

We thank Diane Williams, PhD, from Edanz (https://jp.edanz.com/ac) for editing a draft of this manuscript.

Gill TM, Gahbauer EA, Han L, Allore HG. Trajectories of disability in the last year of life. N Engl J Med. 2010;362:1173–80.
Brown MA, Sampson EL, Jones L, Barron AM. Prognostic indicators of 6-month mortality in elderly people with advanced dementia: a systematic review. Palliat Med. 2013;27:389–400.
Ten Koppel M, Onwuteaka-Philipsen BD, Pasman HR, Bernabei R, Carpenter I, Denkinger MD, et al. Are older long term care residents accurately prognosticated and consequently informed about their prognosis? Results from SHELTER study data in 5 European countries. PLoS One. 2018;13:e0200590.
Erel M, Marcus EL, Dekeyser-Ganz F. Barriers to palliative care for advanced dementia: a scoping review. Ann Palliat Med. 2017;6:365–79.
Makino J, Fujitani S, Twohig B, Krasnica S, Oropello J. End-of-life considerations in the ICU in Japan: ethical and legal perspectives. J Intensive Care. 2014;2:9.
Ministry of Health Labour and Welfare of Japan. About "Guidelines for the decision-making process of end-of-life care" (in Japanese) https://www.mhlw.go.jp/shingi/2007/05/s0521-11.html. Accessed 4 December 2021.
Ministry of Health Labour and Welfare of Japan. Revision of "Guidelines for decision-making process of determining medical treatment and care at the end-of-life". (in Japanese) https://www.mhlw.go.jp/stf/houdou/0000197665.html. Accessed 4 December 2021.
Ouchi Y, Toba K, Ohta K, Kai I, Shimizu T, Higuchi N, et al. Guidelines from the Japan Geriatrics Society for the decision-making processes in medical and long-term care for the elderly: Focusing on the use of artificial hydration and nutrition. Geriatr Gerontol Int. 2018;18:823–7.
Japan Medical Association. Guidelines on End-of-Life Care. Japan Med Assoc J. 2008;51:143–8.
The National Hospice Organization. Medical guidelines for determining prognosis in selected non-cancer diseases. Hosp J. 1996;11:47–63.
Monacelli F, Tafuro M, Molfetta L, Sartini M, Nencioni A, Cea M, et al. Evaluation of prognostic indices in elderly hospitalized patients. Geriatr Gerontol Int. 2017;17:1015–21.
Sugiyama M, Takada K, Shinde M, Matsumoto N, Tanaka K, Kiriya Y, et al. National survey of the prevalence of swallowing difficulty and tube feeding use as well as implementation of swallowing evaluation in long-term care settings in Japan. Geriatr Gerontol Int. 2014;14:577–81.
The Ministry of Education, Culture, Sports, Science and Technology. Ethical Guidelines for Medical and Biological Research Involving Human Subjects. (in Japanese) https://www.lifescience.mext.go.jp/bioethics/seimeikagaku_igaku.html. Accessed 4 December 2021.
Bossuyt PM, Reitsma JB, Bruns DE, Gatsonis CA, Glasziou PP, Irwig L, et al. STARD 2015: an updated list of essential items for reporting diagnostic accuracy studies. BMJ. 2015;351:h5527.
Lunney JR, Lynn J, Hogan C. Profiles of older medicare decedents. J Am Geriatr Soc. 2002;50:1108–12.
Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40:373–83.
Anderson F, Downing GM, Hill J, Casorso L, Lerch N. Palliative performance scale (PPS): a new tool. J Palliat Care. 1996;12:5–11.
Mitchell SL, Miller SC, Teno JM, Kiely DK, Davis RB, Shaffer ML. Prediction of 6-month survival of nursing home residents with advanced dementia using ADEPT vs hospice eligibility guidelines. JAMA. 2010;304:1929–35.
Glas AS, Lijmer JG, Prins MH, Bonsel GJ, Bossuyt PM. The diagnostic odds ratio: a single indicator of test performance. J Clin Epidemiol. 2003;56:1129–35.
Kanda Y. Investigation of the freely-available easy-to-use software 'EZR' (easy R) for medical statistics. Bone Marrow Transplant. 2013;48:452–8.
Hui D, Nooruddin Z, Didwaniya N, Dev R, De La Cruz M, Kim SH, et al. Concepts and definitions for "actively dying," "end of life," "terminally ill," "terminal care," and "transition of care": a systematic review. J Pain Symptom Manage. 2014;47:77–89.
Anzai T, Sato T, Fukumoto Y, Izumi C, Kizawa Y, Koga M, et al. JCS/JHFS 2021 Statement on Palliative Care in Cardiovascular Diseases. Circ J. 2021;85:695–757.
Mitchell SL, Teno JM, Kiely DK, Shaffer ML, Jones RN, Prigerson HG, et al. The clinical course of advanced dementia. N Engl J Med. 2009;361:1529–38.
Kawakami Y, Hamano J. Changes in body mass index, energy intake, and fluid intake over 60 months premortem as prognostic factors in frail elderly: A post-death longitudinal study. Int J Environ Res Public Health. 2020;17:1823.
Arahata M, Oura M, Tomiyama Y, Morikawa N, Fujii H, Minani S, et al. A comprehensive intervention following the clinical pathway of eating and swallowing disorder in the elderly with dementia: historically controlled study. BMC Geriatr. 2017;17:146.
Parikh RB, Manz C, Chivers C, Regli SH, Braun J, Draugelis ME, et al. Machine learning approaches to predict 6-month mortality among patients with cancer. JAMA Netw Open. 2019;2:e1915997.
Seow H, Tanuseputro P, Barbera L, Earle C, Guthrie D, Isenberg S, et al. Development and validation of a prognostic survival model with patient-reported outcomes for patients with cancer. JAMA Netw Open. 2020;3:e201768.
Armstrong MJ, Alliance S, Corsentino P, DeKosky ST, Taylor A. Cause of death and end-of-life experiences in individuals with dementia with Lewy bodies. J Am Geriatr Soc. 2019;67:67–73.
Browne B, Kupeli N, Moore KJ, Sampson EL, Davies N. Defining end of life in dementia: A systematic review. Palliat Med. 2021;35:1733–46.
Smith EE, Ismail Z. Mortality risk models for persons with dementia: A systematic review. J Alzheimers Dis. 2021;80:103–11.
Loizeau AJ, Shaffer ML, Habtemarian DA, Hanson LC, Volandes AE, Mitchell SL. Association of prognostic estimates with burdensome interventions in nursing home residents with advanced dementia. JAMA Intern Med. 2018;178:922–9.
Jimenez G, Tan WS, Virk AK, Low CK, Car J, Ho AHY. Overview of systematic reviews of advance care planning: summary of evidence and global lessons. J Pain Symptom Manage. 2018;56:436–59.e25.
Malhotra C, Sim D, Jaufeerally FR, Hu M, Nadkarni N, Ng CSH, et al. Impact of a formal advance care planning program on end-of-life care for patients with heart failure: results from a randomized controlled trial. J Card Fail. 2020;26:594–8.

Additionalfile1.pdf
File name: Additional_file_1.pdfContents: Text S1: Japanese guidelines for the end-of-life medical decision-making process Table S1: Regular members of the end-of-life case conference Table S2: Baseline characteristics (for diagnostic accuracy analysis)

Accurate Identification and Prognostication of End-of-Life State in Japan Using A Guideline-Based Diagnostic Method: A Diagnostic and Prognostic Study

Status:

Journal Publication

Version 1

Abstract

Figures

Background

Methods

Study Setting

Study Design and Participants

Diagnostic Criteria for EOL State

Data collection and follow-up

Outcomes

Statistical analysis

Results

Study population and baseline characteristics

Associations between EOL-CC diagnosis and primary outcome

Diagnostic odds ratio and other index values of diagnostic accuracy

Discussion

Study limitations

Conclusions

Abbreviations

Declarations

References

Supplementary Files

Status:

Journal Publication

Version 1