Abstract
Objective
To assess the drug prescriptions of nursing home (NH) residents during the 6 months prior to their death, and the impact of the recognition of « life expectancy lower than 6 months » by the NH staff on the prescriptions.
Design
Prospective study.
Setting
175 nursing homes in France.
Participants
6275 residents were included from May to June 2011.
Measurements
The initial drug prescriptions of the residents who deceased within 6 months were compared with those who did not decease. Among the residents deceased within 6 months, the drug prescriptions were compared between the residents who were «considered at the end of their life» and those who were not. Potentially inappropriate prescriptions (PIP) were analyzed using Laroche criteria and a list of therapies considered as inappropriate at the end of life.
Results
498 residents (7.9%) died within 6 months after their inclusion: they had significantly more therapies (8.3 ± 3.8 vs. 7.9 ± 3.5, p=0.048) than non-deceased people. Sixty-one of the residents deceased within 6 months were considered by the NH staff as «end of life residents » (12.2%). They received significantly less drugs (6.4 ± 4.2 vs 8.5 ± 3.6, p<0.001) than NH's residents not identified at the end of their life. They had a more frequent prescription of opioids (p<0.001), and less antipsychotics (p<0.001), lipid-lowering drugs (p=0.006), or antihypertensive therapies (p<0.01). They also received significantly less PIP (59.0% received at least one inappropriate prescription, vs. 87.2%, p<0.001).
Conclusion
An important proportion of nursing home residents received PIP. The quality of prescriptions in patients identified at the end of their life seems to improve, but more than half still receive inappropriate drugs. Special attention in prescribing should be given to these patients presenting a high risk of adverse events.
Key words: Elderly, nursing home, end of life, drug
Introduction
Nursing homes (NH) are designed to receive persons who are often very old, dependent and with multiple pa-thologies. Therefore, the mortality rate in this setting is very high. In USA, 24% of American die in NH and this proportion is expected to reach 40% in 2020 (1). In France, 12% of all deaths occur in NHs, concerning more than 90,000 residents every year (2).
Despite a high number of deaths, there is evidence that NH residents are not receiving optimal palliative or end-of-life care (3). Many barriers have been suggested to explain this gap, including a lack of established palliative care procedures, the limited use of advance care planning but also the difficulty to estimate life expectancy and to recognize the right time to shift from curative to palliative care (4, 5, 6).
Optimal palliative care management involves the optimization of drug prescriptions. One of the first objectives is to manage pain and uncomfortable symptoms. A second objective is to limit adverse drug events, which can be source of an alteration of resident's quality of life, hospitalization and death (7, 8, 9). Adverse drug events may be linked to an alteration of kidney and/or liver function, malnutrition, comorbidities, as well as polypharmacy and potentially inappropriate prescriptions (PIP).
Currently, many tools allow for the evaluation of drug prescriptions in elderly subjects who are not at the end of their life (10, 11, 12). However, no tool was validated to assess the prescriptions of end-of-life patients. Several authors tried to define the principles for appropriate prescribing in end-of-life patients (13, 14, 15). The main objec-tives are (i) to avoid the use of multiple drugs (ii) to focus the prescriptions on the patient's most uncomfortable symptoms (iii) to limit the therapies that may extend his life or that can be used in primary or secondary care if they do not provide short term benefits. The therapeutic objectives should be reassessed on a regular basis, de-pending on his life expectancy, clinical condition, personal choices and beliefs. Finally, very limited research has been done on drug care for end-of-life subjects in NHs (13).
Our work included three objectives. Our initial goal was (i) to compare drug prescriptions (including PIP) of the NH residents deceased within the 6 follow-up months to that of residents non-deceased within 6 months. Our hypothesis was that the residents deceased within 6 months had a specific, but not optimal drug prescription considering their life expectancy. We also aimed to (ii) assess the capacity of the care team in identifying resi-dents having a life expectancy below 6 months and (iii) to analyze whether this prediction of a death within 6 months was related to a change in drug prescription.
Subjects and methods
Study procedures and population
This study was based on the data of the IQUARE study, which is a controlled nonrandomized intervention study, of educational and professional support, that took place in the NHs of the Midi-Pyrénées region, South-Western France. The methodology (16) and results (17) of this study were previously published. Briefly, the main objective of this study was to improve NH quality indicators and to reduce the functional decline of resi-dents with 18-month follow-up. All the NHs in the Midi-Pyrénées region were invited to participate in the study (n=423): those that volunteered and were not participating in another interventional study were eligible. As for the residents, those who lived for at least 30 days at the NH and who did not refuse to take part in the study were eligible. The NHs were split into two intervention groups: “mild and “strong. In the “mild inter-vention group, NHs received their descriptive statistics (structure, health condition of their residents and quality indicators) and those of the regional NHs (to observe how they rate on these indicators compared to other NHs). In the “strong intervention group, the NHs received the same information, but the data were discussed with a hospital geriatrician who suggested improvement strategies for each indicator. In total, 175 NHs and 6275 resi-dents were included in the IQUARE study (Figure 1).
Figure 1.
Flow chart of the IQUARE study
Data collection
Information concerning the NH organization and residents' characteristics were collected online, on a dedicated website, at baseline. The coordinating physician or nurse completed the questionnaire concerning resident's healthcare. Resident's information included: age, gender, comorbidities (using the Charlson comorbidity index), and medications. The degree of dependence was defined by the Groupe Iso Ressource (GIR), a french disability index. The GIR score is calculated according to the degree of physical and mental autonomy and range from 1 (severe disability) to 6 (total independency). Furthermore, the data collection also included a pain assessment using the question: “Does the resident's record include a pain measurement result obtained using a scale? and “Does the resident complain of any pain?. Another question concerned the end of life: “Is your resident facing the final stages of life? (defined as an estimated life expectancy of 6 months or less). This evaluation should express the consensual estimate of the NH staff (nurses, doctors, assistant nurses). Finally, in the event of death during the follow-up, the date of death was recorded for each resident.
Outcome measure
Definition of potentially inappropriate prescriptions (PIP): Inappropriate drug prescriptions were defined according to the principles developed by O'Mahony and Cruz-Jentoft (13, 14, 15). Therefore, the following therapies are considered as inappropriate:
1) The drug considered to be potentially inappropriate with elderly subjects in general. For this purpose, we used the Laroche criteria (11), which are a French adaptation of De Beers criteria (we excluded the use of diazepam and scopolamine that can be appropriately used at the end of life).
2) The drug considered to be inappropriate “specifically in end of life residents: these are mainly the therapies used in primary and secondary prevention for which the expected benefit time frame is lower than life expectancy. The list of these inappropriate therapies includes: lipid-lowering drugs, drugs used to prevent bone fractures and treat osteoporosis, Angiotensin Converting Enzyme (ACE) inhibitors, angio-tensin receptor blockers (ARB), antiplatelet drugs, anti-dementia drugs (cholinesterase inhibitors and memantine), vitamins and minerals.
Statistical analysis
The data were analysed using using STATA version 11.0
1. Resident's characteristics and drug prescriptions were compared using the chi-square test for qualitative varia-bles and the Student-t test for quantitative variables:
- in the overall population according to the deceased or nondeceased status within six months
- in the group of deceased residents according to their “end of life status
2. The care team's ability to identify end of life residents (e.g having a life expectancy of 6 months or less) was studied in terms of sensitivity and specificity. Sensitivity was defined as the probability for the NH staff to be able to identify residents as “end of life resident, if they died within 6 months and the specificity as the proba-bility to consider a resident “not at the end of life if they did not die within 6 months.
Results
In total, from the 6275 residents included in the study: 498 residents (7.9%) died during the first 6 months of the study. These residents were significantly older than those who did not die, were predominantly male, were more dependent according to their GIR score, and had more comorbidities according to the Charlson score (Table 1). They had more frequently a medical history of myocardial infarction, heart failure, dementia, chronic lung disease, kidney disease and cancer. From the 498 deceased residents, those initially identified as “end of life resident are more dependent and have a higher comorbidity score.
Table 1.
Baseline characteristics of residents deceased within 6 months compared to non-deceased ones (n=6275), and of residents deceased within 6 months according to their end of life status
| Overall population (n=6275) | Residents deceased within 6 months (n=498) | |||||
|---|---|---|---|---|---|---|
| Deceased within 6 months (n=498) | Non deceased within 6 months (n=5777) | p | Identified at the end of their life (n=61) | Not identified at the end of their life (n=437) | p | |
| Age, years, mean (SD) | 88.8 (± 6.9) | 85.7 (±8.2) | <0.001 | 87.7 (± 0.9) | 88.9 (±0.3) | 0.178 |
| Gender (Female), n (%) | 346 (69.5 | 4276 (74.0) | 0.027 | 45 (73.8) | 301 (68.9) | 0.437 |
| GIR, mean (SD) | 2.2 (±1.2) | 2.8 (±1.4) | <0.001 | 1.4 (±0.6) | 2.3 (±1.2) | <0.001 |
| MEDICAL HISTORY/ COMORBIDITIES, n (%) | ||||||
| Myocardial infarction | 51 (10.2) | 421 (7.3) | 0.016 | 3 (5.0) | 48 (11.0) | 0.143 |
| Peripheral vascular disease | 94 (18.9) | 1028 (17.8) | 0.546 | 6 (9.8) | 88 (20.1) | 0.054 |
| Heart failure | 140 (28.1) | 1079 (18.7) | <0.001 | 15 (24.6) | 125 (28.6) | 0.514 |
| Hypertension | 263 (52.8) | 3064 (53.0) | 0.923 | 25 (41.0) | 238 (54.5) | 0.048 |
| Hemiplegia | 22 (4.4) | 235 (4.1) | 0.705 | 5 (8.2) | 17 (3.9) | 0.125 |
| Stroke | 68 (13.6) | 727 (12.6) | 0.491 | 10 (16.4) | 58 (13.3) | 0.506 |
| Dementia | 240 (48.6) | 2448 (42.9) | 0.014 | 36 (60.0) | 204 (47.0) | 0.059 |
| Depression | 164 (34.1) | 1981 (34.6) | 0.814 | 15 (24.6) | 152 (35.4) | 0.095 |
| Diabetes | 81 (16.3) | 911 (15.8) | 0.771 | 9 (14.8) | 72 (16.5) | 0.315 |
| Chronic lung disease | 77 (15.5) | 568 (9.8) | <0.001 | 11 (18.0) | 66 (15.1) | 0.553 |
| Gastro-duodenal ulcer | 21 (4.2) | 306 (5.3) | 0.298 | 3 (4.9) | 18 (4.1) | 0.771 |
| Hepatic disease (mild to severe) | 10 (2.0) | 67 (1.2) | 0.099 | 4 (6.6) | 6 (1.4) | 0.007 |
| Kidney disease (mild to severe) | 88 (17.7) | 629 (10.9) | <0.001 | 13 (21.3) | 75 (17.2) | 0.426 |
| Cancer/ hematologic malignancies | 99 (19.9) | 700 (12.1) | <0.001 | 13 (21.3) | 86 (19.7) | 0.765 |
| Index of Charlson comorbidities, mean (SD) | 2.7 (±2.1) | 2.0 (±1.7) | <0.001 | 3.3 (±2.7) | 2.6 (±1.9) | 0.008 |
| PAIN / END OF LIFE, n (%) | ||||||
| Pain assessment in the chart | 90 (18.1) | 620 (10.8) | <0.001 | 27 (44.3) | 63 (14.4) | <0.001 |
| The resident complains of pain | 146 (29.8) | 1321 (23.1) | 0.001 | 29 (47.5) | 117 (27.3) | <0.001 |
| Is the patient at the end of his/her life? | 61 (12.4) | 89 (1.5) | <0.001 | - | - | - |
SD, Standard Derivation
From among the residents who died within 6 months, 12.4% were considered being “at the end of life at base-line by the NH staff (as opposed to 1.5% among the non-deceased residents, p<0.001). Therefore, the results of the question – “Is your resident facing the final stages of life? are quite poor: the question's sensitivity is 12.4% (IC at 95%: 9.6-15.6), its specificity is 98.4% (IC at 95%: 98.1-98.7). The positive predictive value is 40.6% and negative predictive value is 7%.
Table 2 compares the main drug prescriptions of the residents deceased or not within 6 months, and of the 498 residents deceased during the 6 first months according to their “end-of-life status. For the 498 residents de-ceased within 6 months, the main drug categories prescribed were: laxatives (49.8%), antiulcer drugs (42.6%), diuretics (39.2%), paracetamol (36.3%), and benzodiazepines (35.1%). They received significantly more step 3 opioids, antiarrhythmic drugs, heparins, calcium blockers, diuretics, insulin, corticosteroids and anti-gout thera-py but less antidepressants and oral anti-diabetic drugs, than the residents non deceased within 6 months. The prescription of step 3 opioids is significantly more important in the sample of residents identified as being at the end of their life, and these residents also received significantly less antidepressant, antipsychotic, cardiovas-cular, oral anti-diabetic and anti-gout therapies.
Table 2.
Main drug categories prescribed for the 6275 residents of the IQUARE study at baseline, and for the 498 residents deceased within 6 months
| Overall population (n=6275) | Residents deceased within 6 months (n=498) | |||||
|---|---|---|---|---|---|---|
| Deceased within 6 months (n=498) | Non deceased within 6 months (n=5777) | p | Identified at the end of their life (n=61) | Not identified at the end of their life (n=437) | p | |
| ANALGESICS, n (%) | ||||||
| Paracetamol (alone) | 181 (36.3) | 2277 (39.4) | 0.178 | 17 (27.87) | 164 (37.5) | 0.142 |
| Step 2 analgesic | 45 (9.0) | 549 (9.5) | 0.733 | 2 (3.3) | 30 (6.9) | 0.285 |
| Step 3 opioids | 42 (8.4) | 189 (3.3) | <0.001 | 20 (32.8) | 28 (6.4) | <0.001 |
| At least one analgesic therapy | 239 (48.0) | 2668 (46.2) | 0.437 | 34 (55.7) | 205 (46.9) | 0.196 |
| PSYCHOTROPIC DRUGS, n (%) | ||||||
| Antidepressants | 190 (38.1) | 2530 (43.8) | 0.015 | 9 (14.7) | 181 (41.4) | <0.001 |
| Benzodiazepines | 175 (35.1) | 2079 (36.0) | 0.705 | 20 (32.8) | 155 (35.5) | 0.681 |
| Antipsychotics | 121 (24.3) | 1411 (24.4) | 0.949 | 7 (11.5) | 114 (26.1) | 0.013 |
| Hypnotics | 120 (24.2) | 1600 (27.7) | 0.084 | 9 (14.7) | 111 (25.4) | 0.069 |
| CARDIOVASCULAR THERAPIES, n (%) | ||||||
| Antiarrhythmic drugs | 88 (17.7) | 585 (10.1) | <0.001 | 7 (11.5) | 81 (18.5) | 0.176 |
| Heparin | 23 (4.6) | 119 (1.9) | <0.001 | 3 (4.9) | 20 (4.6) | 0.753 |
| Vitamin K Antagonist | 76 (15.3) | 757 (13.1) | 0.173 | 4 (6.6) | 72 (16.5) | 0.044 |
| Beta-blockers | 90 (18.1) | 1016 (17.6) | 0.785 | 4 (6.56) | 86 (19.68) | 0.013 |
| Calcium blockers | 73 (14.7) | 1202 (20.8) | 0.001 | 1 (1.6) | 72 (16.5) | 0.002 |
| Diuretics | 195 (39.2) | 1889 (32.7) | 0.003 | 14 (22.9) | 181 (41.4) | 0.006 |
| OTHER THERAPIES, n (%) | ||||||
| Insulin | 63 (12.6) | 486 (8.4) | 0.001 | 7 (11.5) | 55 (12.6) | 0.806 |
| Oral anti-diabetic drugs | 19 (3.8) | 380 (6.6) | 0.015 | 0 (0.0) | 19 (4.3) | 0.097 |
| Antiulcer drugs (PIP/ H2 blockers) | 212 (42.6) | 2239 (38.8) | 0.094 | 23 (37.7) | 189 (43.2) | 0.412 |
| Antiemetic drugs | 29 (5.8) | 303 (5.2) | 0.580 | 3 (4.9) | 1 (0.2) | <0.001 |
| Laxatives | 248 (49.8) | 2620 (45.3) | 0.056 | 29 (47.5) | 219 (50.1) | 0.706 |
| Antibiotics | 29 (5.8) | 271 (4.7) | 0.256 | 3 (4.9) | 26 (5.9) | 0.747 |
| Corticosteroids (systemic use) | 40 (8.0) | 161 (2.8) | <0.001 | 9 (14.7) | 31 (7.1) | 0.039 |
| Anti-gout therapy | 28 (5.6) | 215 (3.7) | 0.035 | 0 (0.0) | 28 (5.6) | 0.042 |
ACEI, Angiotensin-converting-enzyme inhibitor. ARB II, Angiotensin II Receptor blocker. PIP, Proton Pump Inhibitor.
Table 3 describes the number of drugs and the frequency of PIPs among the overall population according their death status, and among the 498 residents deceased within 6 months according to their “end of life status. Res-idents deceased within 6 months received a larger number of drug therapies (p=0.048), than those non deceased within 6 months. They receive less frequently at least one inappropriate therapy in total. In the group of resi-dents deceased within 6 months, those identified at the end o their life received significantly less drugs (p<0.001). The number of PIP is also significantly decreased in this group: the number of residents receiving at least one inappropriate prescription according to Larroche criteria and/ or one therapy considered as specifically inappropriate in end of life situations is significantly lower, compared to residents not identified at the end of their life by the NH team.
Table 3.
Number of drugs and potentially inappropriate prescription of the 6275 residents of the IQUARE study at baseline, and of the 498 residents deceased within 6 months
| Overall population (n=6275) | Residents deceased within 6 months (n=498) | |||||
|---|---|---|---|---|---|---|
| Deceased within 6 months (n=498) | Non deceased within 6 months (n=5777) | p | Identified at the end of their life (n=61) | Not identified at the end of their life (n=437) | p | |
| DRUG THERAPIES | ||||||
| Number of drugs, mean (SD) | 8.3 (3.8) | 7.9 (3.5) | 0.048 | 6.4 (4.2) | 8.5 (3.6) | <0.001 |
| INAPPROPRIATE THERAPIES ACCORDING TO LARROCHE CRITERIA | ||||||
| At least one inappropriate therapy, n (%) | 217 (43.6) | 2697 (47.1) | 0.182 | 18 (29.5) | 199 (45.5) | 0.018 |
| THERAPIES CONSIDERED AS INAPPROPRIATE SPECIFICALLY IN END OF LIFE SITUATIONS | ||||||
| Lipid-lowering drugs | 49 (9.8) | 883 (15.3) | 0.001 | 0 (0.0) | 49 (11.2) | 0.006 |
| Bisphosphonates | 0 (0.0) | 21 (4.2) | 0.080 | |||
| ACEI/ARB II | 145 (29.1) | 1827 (31.6) | 0.247 | 9 (14.7) | 136 (31.1) | 0.008 |
| Antiplatelet drugs | 174 (34.9) | 2059 (35.6) | 0.754 | 12 (19.7) | 162 (37.1) | 0.008 |
| Anti-dementia drugs | 97 (19.5) | 1329 (23.0) | 0.072 | 8 (13.1) | 89 (20.4) | 0.180 |
| Vitamins and minerals | 155 (31.1) | 1728 (29.9) | 0.571 | 13 (21.3) | 142 (32.5) | 0.077 |
| Total (at least 1 inappropriate therapy) | 363 (72.9) | 4433 (76.7) | 0.052 | 29 (47.5) | 334 (76.4) | <0.001 |
| TOTAL (at least 1 inappropriate prescription: Laroche criteria or inappropriate in end of life) | 417 (83.7) | 5066 (87.7) | 0.011 | 36 (59.0) | 381 (87.2) | <0.001 |
SD, Standard Derivation, ACEI, Angiotensin-converting-enzyme inhibitor. ARB II, Angiotensin II Receptor blocker.
Discussion
The results of this study indicate that NH's residents, during their last six months of life, often receive multiple drugs (8.3 ± 3.8 drugs on average). Moreover, 83.7 % of them have at least one PIP. When identified at the end of life, they receive significantly less drugs (6.4 ± 4.2 vs. 8.5 ± 3.6, p<0.001), and the proportion of subjects having at least one PIP is also reduced (59% vs. 87.2%, p<0.001).
To our knowledge, few studies focused on drug prescriptions in NH subjects at the end of life (13). Most authors have directed their works on the care of residents affected by advanced dementia (18–21) or cancer (22–24).
In our study, the average number of therapies received per resident deceased within 6 months is 8.3 ± 3.8. Residents deceased within 6 months have a higher number of drugs than those non deceased within 6 months: they have more frequently a prescription of heparins, antiarrhythmic drugs, diuretics, or insulin, suggesting that there is an acute decompensated pathology that could be the cause of death. Interestingly, in residents identified at the end of their life, the number of therapies is significantly less important, probably due to the interruption of a certain number of unnecessary therapies. Some studies showed that, when patients get closer to dying, the num-ber of therapies tend to increase. In the study of Chen et al. (25), the mean number of drugs went from 6.0, six months prior to the patient's death, to 7.9 right before the death. They show a significant increase in prescriptions of hypertension and pulmonary medication and laxatives. The same tendency occurs in the study of Cur-row et al. (26): they showed an increase of prescriptions for symptomatic reasons (analgesic treatment, antiemet-ic drug, benzodiazepine) and a decrease of prescriptions for comorbidities. In our study, we do not have the possibility to highlight the evolution of prescriptions, since the data collection was carried out at a fixed date, at inclusion, and may not reflect prescriptions right before death. However we observe an increased prescription of symptomatic treatment such as corticosteroids or analgesic treatment, but not of antiemetic drug or laxative.
Our study shows that only 18.1% of the 498 patients deceased within 6 months had a pain assessment scale in their records and 48% received at least one pain treatment (mainly paracetamol and step 3 opioids). These results are similar to other studies that suggest that pain is under-detected and undertreated in NH (27, 28). When residents are assessed as being “at the end of their life, the pain assessment and treatment are significantly more frequent. Actually, the number of prescriptions for step 3 opioids increases significantly (32.8% vs. 6.4%, p<0.001). However, it remains suboptimal considering that almost 50% of the residents have not a pain assess-ment.
A number of studies have examined the PIPs of elderly subjects living in NH, regardless of their status (end-of-life or not). Ruggiero et al. showed that 5 to 55% of all NH residents (in the United States and Canada) have at least one PIP (29). This number may be even higher in NHs in Europe, reaching 18 to 70% (29). Our study shows a higher prevalence : 72.9% of the residents who died within six months had at least one PIP, and this percentage was greater according to their end of life status (87.2% if “not at the end of life vs. 59.0% if “end of life resident, p<0.001). These results may be explained by several reasons. First, our definition of PIPs is different from most of the studies mentioned in the meta-analysis of Ruggiero that used the Beers criteria. We chose the Laroche criteria, which are a French adaptation of Beers criteria, more adapted to the French practice. Moreover, in the definition of PIP, we include a list of “inappropriate medications at the end of life including drug therapies used in primary and secondary prevention: this may increase significantly the prevalence of PIPs in this study. However, some therapies that are usually considered as inappropriate for elderly subjects (long halflife benzodiazepines, such as diazepam, or anticholinergic therapies, such as scopolamine) may find their place in the symptomatic therapies of end of life patients and are considered appropriate in this analysis.
Our results also suggest that, when the residents are identified at the end of their life, there is a significant diminution of PIPs, concerning both the inappropriate therapies according to Larroche criteria (27.9% vs. 48.5%) and the inappropriate therapies specifically in end of life (59% vs. 87.4%). To our knowledge, no other descriptive study has specifically assessed the prevalence of PIPs of end of life NH residents, therefore not allowing for a comparative analysis. One intervention study showed that the intervention of a mobile palliative care team in nursing home results in a significant diminution of the PIPs in the 30 days before death (30).
It is interesting to stress that several preventive therapies, such as lipid-lowering drugs and bisphosphonates, have been completely interrupted in residents identified at the end of life in this study. Our results are consistent with another study in a palliative care unit, showing that no patients admitted in this unit at a terminal phase of disease were prescribed a lipidlowering therapy (31). Another study conducted in 3 nursing homes in Norway, shows that the number of curative/preventive drugs decreased at the end of life but mainly on the day of death (32). Therefore, there seems to be a tendency in interrupting preventive therapies as the end of life approaches.
Finally, the residents recognized at the end of their life are not less treated by vitamin and mineral supplements (21.3% vs. 32.5%, p=0.007). Similarly, they do not receive less antidementia therapies than others (13.1% vs. 20.4%, p=0.180). Several studies showed that the patients with dementia often die under their anti-dementia therapy (33,34). Daeillo et al. have suggested that the brutal interruption of anticholinesterases would relate to a recrudescence of behavioral disorders (35), but only a few studies have actually assessed the impact of the inter-ruption of anti-dementia therapies at the end of life. This also underlines the difficulty which may sometimes occur for physicians as well as for the patient or his/her family in interrupting a long-term therapy.
This study underlines the fact that the care teams underestimate the risk of death within 6 months. Although they come across end of life and death more than anywhere else, the NH staff are not efficient in identifying the residents living their last 6 months of life. Only 12% of the 498 deceased residents during the 6 first months are considered to be at the end of their life upon inclusion. This inability of the teams in identifying the residents living their last months of life could be explained by the complexity of poly-pathological situations, with an unpredictable course (36). This is definitely regrettable, because our work shows that the identification of end of life subjects improves residents' care. This requires the more systematic use of tools for the assessment of life expectancy of NH residents (37) or the expertise of a palliative care team in improving the relevance of care pro-jects.
This study has several limitations. First, we have assumed that the residents deceased within 6 months had a high risk of mortality and that they should have received palliative care and therefore an adaptation of their drug prescription. However, we do not know the patients' causes of death: many of them may have died due to an acute pathology. There is also the question of defining the “end of life in this paper, which relies on the re-sponse of the physician and of the care team to the question: “Is the patient at the end of his/her life?. Identifying the end of life with polypathological elderly subjects, whose trajectories are often unpredictable, remains a complex task (38). Another limitation is the collection of data concerning drug therapies, which is performed in the beginning of the study: we do not know the potential changes occurred between the beginning of the study and the resident's death. Several changes may have occurred in the meantime, which we were not able to consider.
Conclusion
Our study highlights the fact that the elderly subjects in nursing homes during their last months of life often receive multiple drugs and many of them have inappropriate prescriptions. A small proportion of the end-of-life residents are identified as living their last months of life. When they are assessed as being at the end of their life, the number of drugs is reduced and the quality of prescriptions improves. However, efforts should be made for pain management and for the reassessment of end-of-life therapies. Intervention studies seem to be necessary in order to assess and improve the quality of drug prescriptions for the terminally-ill patients in nursing homes.
Acknowledgements
The authors would like to thank the members of the IQUARE Research Group (members of COPIL: Dr. Jean-Jacques Morfoisse, Gwenaelle Buatois, Dr. Catherine Marchal, Pascal Degauque, and Sabine Pi) and the 57 members of the Technical Committee for their valuable work. We also thank all those who made this research possible, particularly hospital geriatricians and coordinating physi-cians, nurses and directors of participating NHs, and staff from the Agence Regional de la Santé— Midi-Pyrenées (ARS-MP) who contributed to this study. We also thank the ARS-MP for funding.
Ethics approval and consent to participate
The study protocol was approved by the ethic committee of the Toulouse University Hospital (CHU Toulouse) and the Consultative Committee for the Treatment of Research Information on Health (Commission Nationale de l'Informatique et des Libertés - CNIL (07-438), and follow the principles of the 1964 Helskinki declaration.
Conflict of interest
The authors have no conflict of interest to declare.
Funding
This work was supported by the Gerontopole of Toulouse and the Regional Health Agency Midi-Pyrénées (ARS-MP). The funders had no role in data analysis, interpretation of the data, preparation of the man-uscript and decision to publish.
Author's contribution
YR and BV conceived and designed the study. SS performed the statistical analysis. CR and SS designed and led the literature search. CR and SS wrote the first draft of the paper and prepared figures and tables. PSB, FN, YR and BV revised the manuscript and approved the final version.
References
- 1.Christopher M, Benchmarks to improve end of life care, 2000
- 2.Blanpain N, Chardon O. Projections de population à l’horizon 2060: “Un tiers de la population âgé de plus de 60 ans. Insee Prem. 2010;1320:1–4. [Google Scholar]
- 3.Wallace CL, Swagerty D, Barbagallo M, Vellas B, Cha HB, Holmerova I, et al. IAGG/IAGG GARN International Survey of End-of-Life Care in Nursing Homes. J Am Med Dir Assoc. 2017;18:465–469. doi: 10.1016/j.jamda.2017.04.002. 10.1016/j.jamda.2017.04.002 PubMed PMID: 28549702. [DOI] [PubMed] [Google Scholar]
- 4.Swagerty D. Integrating palliative care in the nursing home: an interprofessional opportunity. J Am Med Dir Assoc. 2014;15:863–865. doi: 10.1016/j.jamda.2014.09.008. 10.1016/j.jamda.2014.09.008 PubMed PMID: 25441100. [DOI] [PubMed] [Google Scholar]
- 5.Weathers E, O’Caoimh R, Cornally N, Fitzgerald C, Kearns T, Coffey A, et al. Advance care plan-ning: A systematic review of randomised controlled trials conducted with older adults. Maturitas. 2016;91:101–109. doi: 10.1016/j.maturitas.2016.06.016. 10.1016/j.maturitas.2016.06.016 PubMed PMID: 27451328. [DOI] [PubMed] [Google Scholar]
- 6.Oliver DP, Porock D, Zweig S. End-of-life care in U.S. nursing homes: a review of the evidence. J Am Med Dir Assoc. 2004;5:147–155. doi: 10.1097/01.JAM.0000123063.79715.8E. PubMed PMID: 15115574. [DOI] [PubMed] [Google Scholar]
- 7.Perri M, Menon AM, Deshpande AD, Shinde SB, Jiang R, Cooper JW, et al. Adverse outcomes associated with inappropriate drug use in nursing homes. Ann Pharmacother. 2005;39:405–411. doi: 10.1345/aph.1E230. 10.1345/aph.1E230 PubMed PMID: 15671088. [DOI] [PubMed] [Google Scholar]
- 8.Ruggiero C, Dell’Aquila G, Gasperini B, Onder G, Lattanzio F, Volpato S, et al. Potentially inap-propriate drug prescriptions and risk of hospitalization among older, Italian, nursing home residents: the ULISSE project. Drugs Aging. 2010;27:747–758. doi: 10.2165/11538240-000000000-00000. 10.2165/11538240-000000000-00000 PubMed PMID: 20809664. [DOI] [PubMed] [Google Scholar]
- 9.Tamura BK, Bell CL, Inaba M, Masaki KH. Outcomes of polypharmacy in nursing home residents. Clin Geriatr Med. 2012;28:217–236. doi: 10.1016/j.cger.2012.01.005. 10.1016/j.cger.2012.01.005 PubMed PMID: 22500540. [DOI] [PubMed] [Google Scholar]
- 10.Beers MH, Ouslander JG, Rollingher I, Reuben DB, Brooks J, Beck JC. Explicit criteria for deter-mining inappropriate medication use in nursing home residents. UCLA Division of Geriatric Medicine. Arch Intern Med. 1991;151:1825–1832. 10.1001/archinte.1991.00400090107019 PubMed PMID: 1888249. [PubMed] [Google Scholar]
- 11.Laroche M-L, Charmes J-P, Merle L. Potentially inappropriate medications in the elderly: a French consensus panel list. Eur J Clin Pharmacol. 2007;63:725–731. doi: 10.1007/s00228-007-0324-2. 10.1007/s00228-007-0324-2 PubMed PMID: 17554532. [DOI] [PubMed] [Google Scholar]
- 12.O’Mahony D, O’Sullivan D, Byrne S, O’Connor MN, Ryan C, Gallagher P. STOPP/START crite-ria for potentially inappropriate prescribing in older people: version 2. Age Ageing. 2015;44:213–218. doi: 10.1093/ageing/afu145. 10.1093/ageing/afu145 PubMed PMID: 25324330. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Cruz-Jentoft AJ, Boland B, Rexach L. Drug therapy optimization at the end of life. Drugs Aging. 2012;29:511–521. doi: 10.2165/11631740-000000000-00000. 10.2165/11631740-000000000-00000 PubMed PMID: 22642784. [DOI] [PubMed] [Google Scholar]
- 14.Gordon M. Ethical challenges in end-of-life therapies in the elderly. Drugs Aging. 2002;19:321–329. doi: 10.2165/00002512-200219050-00001. 10.2165/00002512-200219050-00001 PubMed PMID: 12093319. [DOI] [PubMed] [Google Scholar]
- 15.O’Mahony D, O’Connor MN. Pharmacotherapy at the end-of-life. Age Ageing. 2011;40:419–422. doi: 10.1093/ageing/afr059. 10.1093/ageing/afr059 PubMed PMID: 21622981. [DOI] [PubMed] [Google Scholar]
- 16.De Souto Barreto P, Lapeyre-Mestre M, Mathieu C, Piau C, Bouget C, Cayla F, et al. A multicen-tric individually-tailored controlled trial of education and professional support to nursing home staff: re-search protocol and baseline data of the IQUARE study. J Nutr Health Aging. 2013;17:173–178. doi: 10.1007/s12603-013-0008-9. 10.1007/s12603-013-0008-9 PubMed PMID: 23364498. [DOI] [PubMed] [Google Scholar]
- 17.Rolland Y, Mathieu C, Piau C, Cayla F, Bouget C, Vellas B, et al. Improving the Quality of Care of Long-Stay Nursing Home Residents in France. J Am Geriatr Soc. 2016;64:193–199. doi: 10.1111/jgs.13874. 10.1111/jgs.13874 PubMed PMID: 26782872. [DOI] [PubMed] [Google Scholar]
- 25.Chen I-C, Liu M-L, Twu F-C, Yuan C-H. Use of medication by nursing home residents nearing end of life: a preliminary report. J Nurs Res JNR. 2010;18:199–205. doi: 10.1097/JNR.0b013e3181ed5731. 10.1097/JNR.0b013e3181ed5731 PubMed PMID: 20808079. [DOI] [PubMed] [Google Scholar]
- 26.Currow DC, Stevenson JP, Abernethy AP, Plummer J, Shelby-James TM. Prescribing in palliative care as death approaches. J Am Geriatr Soc. 2007;55:590–595. doi: 10.1111/j.1532-5415.2007.01124.x. 10.1111/j.1532-5415.2007.01124.x PubMed PMID: 17397439. [DOI] [PubMed] [Google Scholar]
- 27.Hunnicutt JN, Ulbricht CM, Tjia J, Lapane KL. Pain and pharmacologic pain management in long-stay nursing home residents. Pain. 2017;158:1091–1099. doi: 10.1097/j.pain.0000000000000887. 10.1097/j.pain.0000000000000887 PubMed PMID: 28267065, PMCID 5435509. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Takai Y, Yamamoto-Mitani N, Okamoto Y, Koyama K, Honda A. Literature review of pain preva-lence among older residents of nursing homes. Pain Manag Nurs. 2010;11:209–223. doi: 10.1016/j.pmn.2010.08.006. 10.1016/j.pmn.2010.08.006 PubMed PMID: 21095596. [DOI] [PubMed] [Google Scholar]
- 29.Ruggiero C, Lattanzio F, Dell’Aquila G, Gasperini B, Cherubini A. Inappropriate drug prescriptions among older nursing home residents: the Italian perspective. Drugs Aging. 2009;26(1):15–30. doi: 10.2165/11534630-000000000-00000. 10.2165/11534630-000000000-00000 PubMed PMID: 20136166. [DOI] [PubMed] [Google Scholar]
- 30.Suhrie EM, Hanlon JT, Jaffe EJ, Sevick MA, Ruby CM, Aspinall SL. Impact of a geriatric nursing home palliative care service on unnecessary medication prescribing. Am J Geriatr Pharmacother. 2009;7:20–25. doi: 10.1016/j.amjopharm.2009.02.001. 10.1016/j.amjopharm.2009.02.001 PubMed PMID: 19281937, PMCID 4870890. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Russell BJ, Rowett D, Abernethy AP, Currow DC. Prescribing for comorbid disease in a palliative population: focus on the use of lipid-lowering medications. Intern Med J. 2014;44:177–184. doi: 10.1111/imj.12340. 10.1111/imj.12340 PubMed PMID: 24341863. [DOI] [PubMed] [Google Scholar]
- 32.Jansen K, Schaufel MA, Ruths S. Drug treatment at the end of life: an epidemiologic study in nurs-ing homes. Scand J Prim Health Care. 2014;32:187–192. doi: 10.3109/02813432.2014.972068. 10.3109/02813432.2014.972068 PubMed PMID: 25363144, PMCID 4278400. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Daiello LA, Ott BR, Lapane KL, Reinert SE, Machan JT, Dore DD. Effect of discontinuing cholin-esterase inhibitor therapy on behavioral and mood symptoms in nursing home patients with dementia. Am J Geriatr Pharmacother. 2009;7:74–83. doi: 10.1016/j.amjopharm.2009.04.002. 10.1016/j.amjopharm.2009.04.002 PubMed PMID: 19447360. [DOI] [PubMed] [Google Scholar]
- 36.Murray S, Kendall M, Boyd K, Sheikh A. Illness trajectories and palliative care. BMJ. 2005;330:1007–1011. doi: 10.1136/bmj.330.7498.1007. 10.1136/bmj.330.7498.1007 PubMed PMID: 15860828, PMCID 557152. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Yourman LC, Lee SJ, Schonberg MA, Widera EW, Smith AK. Prognostic indices for older adults: a systematic review. JAMA. 2012;307:182–192. doi: 10.1001/jama.2011.1966. 10.1001/jama.2011.1966 PubMed PMID: 22235089, PMCID 3792853. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Thomas K, THe Gold Standart Framework, 2011
Uncited references
- 18.Agar M, Beattie E, Luckett T, Phillips J, Luscombe G, Goodall S, et al. Pragmatic cluster random-ised controlled trial of facilitated family case conferencing compared with usual care for improving end of life care and outcomes in nursing home residents with advanced dementia and their families: the IDEAL study protocol. BMC Palliat Care. 2015;14:63. doi: 10.1186/s12904-015-0061-8. 10.1186/s12904-015-0061-8 PubMed PMID: 26589957, PMCID 4654825. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Husebo BS, Achterberg W, Flo E. Identifying and Managing Pain in People with Alzheimer’s Dis-ease and Other Types of Dementia: A Systematic Review. CNS Drugs. 2016;30:481–497. doi: 10.1007/s40263-016-0342-7. 10.1007/s40263-016-0342-7 PubMed PMID: 27240869, PMCID 4920848. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Reinhardt JP, Boerner K, Downes D. The Positive Association of End-of-Life Treatment Discus-sions and Care Satisfaction in the Nursing Home. J Soc Work End—Life Palliat Care. 2015;11:307–322. doi: 10.1080/15524256.2015.1107805. 10.1080/15524256.2015.1107805 PubMed PMID: 26654063. [DOI] [PubMed] [Google Scholar]
- 21.Van Soest-Poortvliet MC, van der Steen JT, Gutschow G, Deliens L, Onwuteaka-Philipsen B d, Vet HCW, et al. Advance Care Planning in Nursing Home Patients With Dementia: A Qualitative Inter-view Study Among Family and Professional Caregivers. J Am Med Dir Assoc. 2015;16:979–989. doi: 10.1016/j.jamda.2015.06.015. 10.1016/j.jamda.2015.06.015 PubMed PMID: 26255099. [DOI] [PubMed] [Google Scholar]
- 22.Bainbridge D, Seow H, Sussman J, Pond G. Factors associated with acute care use among nursing home residents dying of cancer: a population-based study. Int J Palliat Nurs. 2015;21:349–356. doi: 10.12968/ijpn.2015.21.7.349. 10.12968/ijpn.2015.21.7.349 PubMed PMID: 26203955. [DOI] [PubMed] [Google Scholar]
- 23.Rodin MB. Cancer patients admitted to nursing homes: what do we know. J Am Med Dir Assoc. 2008;9:149–156. doi: 10.1016/j.jamda.2007.11.011. 10.1016/j.jamda.2007.11.011 PubMed PMID: 18294596. [DOI] [PubMed] [Google Scholar]
- 24.Monroe TB, Carter MA, Feldt KS, Dietrich MS, Cowan RL. Pain and hospice care in nursing home residents with dementia and terminal cancer. Geriatr Gerontol Int. 2013;13:1018–1025. doi: 10.1111/ggi.12049. 10.1111/ggi.12049 PubMed PMID: 23506621, PMCID 4850827. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Weschules DJ, Maxwell TL, Shega JW. Acetylcholinesterase inhibitor and N-methyl-D-aspartic acid receptor antagonist use among hospice enrollees with a primary diagnosis of dementia. J Palliat Med. 2008;11:738–745. doi: 10.1089/jpm.2007.0125. 10.1089/jpm.2007.0125 PubMed PMID: 18588406. [DOI] [PubMed] [Google Scholar]
- 34.Herrmann N, Gill SS, Bell CM, Anderson GM, Bronskill SE, Shulman KI, et al. A population-based study of cholinesterase inhibitor use for dementia. J Am Geriatr Soc. 2007;55:1517–1523. doi: 10.1111/j.1532-5415.2007.01377.x. 10.1111/j.1532-5415.2007.01377.x PubMed PMID: 17697100. [DOI] [PubMed] [Google Scholar]