Systemic factors and mortality in elderly patients with pressure ulcers

Abstract

The aim of this article was to identify specific systemic factors associated with existence of pressure ulcers (PUs) and the effect on survival from the time of admission. Patients admitted to the Skilled Nursing Department of the Herzog Hospital, Jerusalem, between 1 July 2008 and 31 December 2011. Of the 174 admitted patients (mean age: 77·4 ± 13·2 years), 107 (61·5%) had pre‐existing PUs and 67 (38·5%) did not have PUs. Major systemic factors were assessed for each patient at the time of admission: sociodemographic characteristics, comorbidities, use of urinary catheter, tube feeding and tracheostomy; nutritional state; Global Deterioration Scale, Glasgow Coma Scale and Norton Scale. Complications such as the number of provided antibiotic courses, and length and outcomes of hospitalisation were identified at the end of the study. In the univariate analysis, patients in the PU group had significantly prevalent characteristics including advanced age, low cognitive and consciousness function, low Norton scale, Parkinson's disease and anaemia due to chronic diseases, low nutritional parameters and higher number of antibiotics provided. Conditions that were associated with PUs in multiple regression analyses included advanced dementia (OR = 3·0, 95% CI: 1·4–6·3; P = 0·002), urinary catheter usage (OR = 2·25, 95% CI: 1·06–4·7; P = 0·03), low body mass index, BMI (OR = 0·92, 95% CI: 0·86–0·99; P = 0·02) and anaemia level (OR = 0·7, 95% CI: 0·58–0·9; P = 0·004). The median survival time of patients with PUs was significantly lower than the non PUs group (94 versus 414 days, respectively) (P = 0·005, log rank test). Length of stay was also significantly lower in the PU group (166 versus 270 days, P = 0·02). The existence of PUs may indicate a final common pathway of various systemic factors (geriatric conditions, diseases and frailty dysfunction).

Introduction

Pressure ulcers (PUs) are a major problem confronting the frail elderly in all medical facilities 1. With the prolonged life expectancy and increase in the number of disabled and mobility challenged ageing population, the PU phenomenon will increase 2. It has been noted that the prognosis of PU is highly variableand sometimes life threatening 3, 4.

The approach to the treatment of all PU is generally similar and concentrates on local wound treatment; the question arises which factors influence the development of the wounds and the prognosis of the patients. The following questions may arise during treatment of PU in an elderly patient: Is the development of PU part of the ageing process indicative of overall frailty and medical deterioration? Is PU predictive of impending death? Or is it limited to local skin condition?

We hypothesised that PU is the final common pathway of the ageing process plus comorbidity resulting in frailty, and the final outcome is geriatric syndrome. Geriatric syndrome is a clinical condition resulting from accumulation and interaction of multiple local and systemic risk factors that impair various systems and organs with a final common pathway resulting in PU syndrome 5.

Inouye et al. reviewing the medical literature from January 1990 through December 2005 identified several PU risk factors 6. Four of the primary PU risk factors included older age, cognitive impairment, functional impairment and impaired mobility, all uniquely common in other geriatric syndromes including falls, delirium and urinary incontinence 7.

PU in the elderly patients requires a management policy different from the approaches treating younger patients with spinal cord injury 8. The treatment of this older age group should involve a comprehensive approach and take into account related factors including age‐related disorder, comorbidity and disability, and not focus purely on localised therapy 9.

The aim of this study was to identify specific systemic factors associated with existence of a PU during hospitalisation and to assess the association of pre‐existing PU on survival.

Methods

All data reviewed were from patients hospitalised in the Skilled Geriatric Nursing Department of the Herzog Hospital in Jerusalem between 1 July 2008 and 31 December 2011.

Study inclusion criteria included patients admitted consecutively to the Skilled Geriatric Nursing Department during the study duration. Admission criteria included patients bound to bed and additionally having one or more of the categories: extensive and deep PU (grades III and IV), terminal cancer requiring palliative care, renal failure requiring haemodialysis, oxygen dependence, non invasive ventilation [Continuous Positive Airway Pressure (CPAP) use] and tracheostomy. Study exclusion criteria were children (<18 years of age) and pregnant women.

The patient workup upon admission included assessment by diverse discipline: nurse for skin examination and the assessment of Norton Scale, the attending physician performed physical examination including the Glasgow Coma Scale (GCS) assessment and routine blood analysis and a dietitian assessed the weight, body mass index (BMI) and recommended calorie and protein intake. The physiotherapist and occupational therapist evaluated the muscle tonus and cognitive state, and speech therapist the swallowing capacity. The social worker assessed the emotional and economic resources of the family.

Follow‐up periodic assessments were performed weekly for the group with PU, including the assessment of PU site, appearance, grade and size, with treatment decisions in a team approach by the physician and nurse. Among those patients without PUs, a routine weekly bedside examination was made by the treating physician.

For every patient, a structured questionnaire was completed by the attending physician upon admission and at the end of follow‐up (date of discharge, death or the end of study period on 31 December 2011). Data were collected from the medical, nursing, social and nutritional files and included sociodemographic characteristics (age, gender, family status and number of children), comorbidities, use of urinary catheter, tube feeding, tracheostomy and PU assessment. Mental state assessment (e.g. the Mini‐Mental State Examination) was impossible to perform in most cases. Therefore, we used the Reisberg's Global Deterioration Scale (GDS) stages of dementia of Alzheimer type (SDAT) 10, the GCS and the Norton Scale. According to the GDS Scale, stages of cognitive level are as follows: stage 1: no cognitive decline; stage 2: very mild cognitive decline; stage 3: mild cognitive decline; stage 4: moderate cognitive decline; stage 5: moderately severe cognitive decline; stage 6: severe cognitive decline and stage 7: very severe cognitive decline. The GCS measures consciousness [score ranging 3–15, with a score of 5–9 indicating vegetative state, whereas >10 indicates alertness 11]. The Norton Scale score ranges from 4 to 24 points. It assesses five risk‐based items: sensory perception, physical and mental condition, incontinence, activities and mobility. The level considered as high risk for developing PUs is <14 points. Data on nutritional status included BMI, weight, cholesterol and total lymphocyte count. Serum albumin, total protein and haemoglobin (Hb) levels are also included. Anaemia of chronic disease is defined as normocytic and normochromic Hb.

Complications such as systemic infections expressed by number of antibiotic courses provided, length and outcomes of hospitalisation were identified at the end of the study. The study population was divided into patients with existing PU upon admission and those without PU. Both groups were followed up during their stay in the department until death, discharge or the end of the study on 31 December 2011. There were no losses to follow‐up or drop‐outs from the follow‐up group. The study was approved by the Helsinki Committee Herzog Hospital.

Statistical methods

Descriptive statistics, by PU (yes/no) at study initiation, are presented for all study participants. Logistic regression models were used to estimate the odds ratio (OR) and 95% confidence intervals (CIs) for PU by study covariates. A multivariate logistic regression model was fitted for a selected group of variables at baseline, which were dropped from the model by backward elimination until all factors were significant at α = 0·10. Variables entered included age, gender, Norton scale, albumin, Hb, BMI, GDS, GCS, urinary catheter usage, anaemia of chronic disease, Parkinson's disease and antibiotic treatment. All variables were entered at the first stage and backward elimination was used to remove non significant factors.

The association between PU and survival time was assessed by the Kaplan–Meier model. We included all patients in the cross‐sectional analyses; however, in the mortality and survival analyses we excluded 19 patients who were admitted before study onset (1 July 2008) because we could not account for the time of admission through the study onset.

All data were analysed with SPSS software (version 17.0; SPSS, Chicago, IL). All statistical tests were two‐sided. A P value below 0·05 was considered significant.

Results

A total of 174 patients were included in this study; of these, 107 (61·5%) were with PUs at study initiation and 67 (38·5%) without PUs. The mean age of the entire group was 77·4 ± 13·2 years (median age 81).

The characteristics of the two groups are compared in univariate analysis (Table 1). The PU group was characterised by a significantly older population (79·24 ± 12 versus 74·61 ± 14 years, P = 0·03). Male and married individuals in the PU group were in higher proportion but did not reach significantly (63% versus 37% P = 0·64). Two‐thirds (115 of 174) of the entire study population were transferred from general hospitals after acute episodes of illness, with 57% of the PU group versus 43%, P = 0·05.

Table 1.

Characteristics of study population^*

Characteristics	Pressure ulcer (n = 107)	Non pressure ulcer (n = 67)	OR	95% CI	P value
Age (years) (mean ± SD)	79·24 ± 12·92	74·61 ± 14·55	1·02	1·00–1·04	0·03*
Gender (male)	55 (63)	32 (37)	1·15	0·62–2·13	0·64
Transfer from general hospital	65 (57)	50 (43)	0·51	0·26–1·0	0·05
Advanced dementia	63 (68)	30 (32)	1·76	0·95–3·27	0·07
Cerebrovascular accident	40 (57)	30 (43)	0·73	0·39–1·37	0·33
Parkinson's disease	17 (94·5)	1(5·5)	12·5	1·61–95·94	0·01*
Diabetes mellitus	40 (55·5)	32 (44·5)	0·65	0·36–1·21	0·17
Ischaemic heart disease	52 (62·5)	31 (37·5)	1·09	0·59–2·02	0·76
Anaemia chronic disease	55 (71·5)	22 (28·5)	2·16	1·14–4·08	0·01*
Peripheral vascular disease	15 (71·5)	6 (28·5)	1·65	0·60–4·5	0·32
Global Deterioration Scale (GDS) total score, mean†	6·52 ± 0·56	6·30 ± 0·74	1·73	1·07–2·81	0·02*
Glasgow Coma Scale (GCS) total score, mean†	11·00 ± 3·16	12·22 ± 2·97	0·87	0·78–0·97	0·01*
Total Norton Scale, mean†	8·75 ± 2·05	10·07 ± 3·0	0·81	0·71–0·92	0·001*
Urinary catheter usage	73 (66)	37 (34)	1·74	0·92–3·27	0·08
Tracheostomy	18 (58)	13 (42)	0·84	0·38–1·85	0·66
Tube feeding	81 (66·4)	41 (33·6)	1·13	0·87–1·45	0·34
Number of antibiotics provided†	54 (67·5)	26 (32·5)	2·26	1·22–4·20	0·001*
Length stay (days), mean	166 ± 247	270 ± 292	0·99	0·997–1	0·02*

CI, confidence interval; OR, odds ratio.

^*Significant at P < 0·05. values are in (%).

^†Global Deterioration Scale (GDS), score range: 1 (no disability)–7 (severe dementia); Glasgow Coma Scale (GCS), total range 3–15; score 5–9 indicates vegetative state, >10 alert; Norton Scale, score range 4–24. The number of antibiotics provided during hospitalisation indicates the level of infection.

Underlying medical conditions were significantly higher in the PU group, including Parkinson's disease and anaemia of chronic disease (Parkinson's disease: 94·5% versus 5·5%, P = 0·01, OR:12·5, 95% CI: 1·61–95·9; anaemia 71·5% versus 28·5%, P = 0·01, OR: 2·16, 95% CI: 1·14–4·08). There was a trend towards significantly different rates of advanced dementia between the two groups: 68% versus 32%, P = 0·07, OR: 1·76, 95% CI: 0·95–3·27.

The prevalence of other coexisting conditions such as cerebrovascular accidents, diabetes mellitus, ischemic heart disease and peripheral vascular disease was not significantly different between the two groups (Table 1). The mean GDS score was significantly higher in the PU study group (6·52 versus 6·30, P = 0·02). Accordingly, the mean GCS score was significantly lower in the PU study group (11·0 versus 12·0, P = 0·01) and the total Norton score was also significantly lower in the PU group (8·75 ± 2·0 versus 10·07 ± 3·0; P = 0·001).

A trend to significant higher prevalence use of urinary catheter was found in the PU group (66% versus 34%, P = 0·08, OR: 1·74, 95% CI: 0·92–3·27). No difference was found between the groups with regard to the presence of a tracheostomy and/or a feeding tube.

Nutritional (weight and BMI) and non nutritional (Hb levels, serum albumin and total protein) parameters at the time of admission were all significantly lower in the PU group (weight: 61·7 versus 68, P = 0·02; BMI: 22·9 versus 25·5, P = 0·007; Hb: 10·46 versus 11·5, P = 0·0002; albumin: 26·5 versus 30·6, P = 0·0001 and total protein: 62·6 versus 67·0, P = 0·02). There were no differences in total lymphocyte count and total cholesterol in both the groups (Table 2).

Table 2.

Nutritional and non nutritional indicators of the study population

Characteristics	PU	No PU	OR	95% CI	P value
Weight (kg), N	97	59	0·97	0·95–0·99	0·02*
Mean	61·7 ± 13·25	68 ± 19·51
BMI (kg/m2), N	90	59	0·91	0·86–0·97	0·007*
Mean	22·9 ± 5·14	25·5 ± 6·0
Albumin (g), N	105	67	0·89	0·84–0·94	0·0001*
Mean	26·5 ± 6·26	30·6 ± 5·65
Hb (g %), N	105	67	0·67	0·55–0·83	0·0002*
Mean	10·46 ± 1·54	11·5 ± 1·7
Total protein	103	65	0·95	0·92–0·99	0·02*
Mean	62·6 ± 13·5	67 ± 6·6
Cholesterol	97	62	1·01	0·82–1·27	0·87
Mean	3·78 ± 1·60	3·74 ± 1·10
Total lymphocyte count	104	67	1	1·0–1·0	0·56
Mean	1731 ± 783	1659 ± 811

BMI, body mass index; CI, confidence interval; OR, odds ratio; PU, pressure ulcer.

Eighty (46%) patients of both groups received one or more courses of antibiotic treatment. The patients in the PU group had a significantly higher incidence of antibiotic treatment (67·5% versus 32·5%, P = 0·001, OR: 2·26, 95% CI: 1·22–4·20). They also had significantly shorter hospital stay (166 versus 270 days; P = 0·02; Table 1).

In multivariate analysis (Table 3), the factors including advanced dementia (OR: 3·0, 95% CI: 1·4–6·3; P = 0·002), urinary catheter usage (OR: 2·25, 95% CI: 1·06–4·77; P = 0·03), low BMI (<23 kg/m²) and low Hb (<11 g %) at the time of admission were associated with PU group (OR: 0·92, 95% CI: 0·86–0·99, P = 0·02; OR: 0·73, 95% CI: 0·58–0·90, P = 0·004, respectively).

Table 3.

The odds ratios of pressure ulcers according to the multivariate logistic regression

Characteristics	PU (%)	No PU (%)	OR	95% CI	P value
Advance dementia	76	32	3·0	1·44–6·31	0·002
Urinary catheter usage	66	33	2·25	1·06–4·7	0·03
BMI (kg/m2), mean	56	44	0·92	0·86–0·99	0·02
Hb (g %), mean	63	37	0·73	0·58–0·90	0·004

BMI, body mass index; CI, confidence interval; OR, odds ratio; PU, pressure ulcer.

A total of 118 (67·8%) patients died during hospitalisation, consisting of 80 (72%) from the PU group compared with 38 (60%) of the non PU group. The median survival time of patients with PU was significantly lower than the non PU group (94 versus 414 days, respectively; P = 0·005, log rank test) (Figure 1).

Figure 1.

Kaplan–Meier cumulative survival of patients according to pressure ulcer (PU) group and no PU group.

Discussion

This study documented that specific geriatric conditions and certain systemic diseases are highly significant in multivariate and univariate analyses associated with PU existence. These geriatric conditions suggest a need to focus specifically on these conditions for the prevention and treatment of PU. Additionally, this study indicates the existence of PUs as potential accelerant for death.

Highly significant systemic factors in multivariate analysis, including advanced dementia, urinary catheter usage, low BMI and low Hb level, were found to be independent contributing factors in the existence of PU. In the univariate analysis, patients in the PU group were significantly older, with higher prevalent underlying medical conditions; Parkinson's disease and anaemia of chronic disease; high cognitive (GDS) and consciousness (GCS) dysfunction; low nutritional parameters and higher antibiotic treatment.

Our finding that the patients in the PU group were significantly older is in line with eight studies showing age as a risk factor for PU 12, 13, 14, 15, 16, 17, 18. Margolis et al. also found that higher risk for PU increased with age; Nonagenarian's PUs risk was higher than in the age span of 80–85 years [adjusted relative risk (RR) 7·47 versus 5·0] 19.

Transfer from acute department of general hospitals to a skilled nursing unit indicates the presence of acute conditions and worsening of the PU severity to grades III and IV 14, 17, 20, 21, 22. Patients in the PU group suffered more from neurodegenerative condition [advanced dementia, Parkinson's disease, high Reisberg deterioration of cognitive state (GDS) and low consciousness state (GCS)].

Advanced dementia is expressed, in our study, as severe cognitive decline reflected in Reisberg GDS with an elevated mean ranking of 6·52–7, resulting in the existence of PU. Six studies showed that cognitive impairment is a risk factor for PU 12, 13, 18, 19, 20, 23. A prospective cohort study of 323 nursing home residents with advanced dementia (in a follow‐up of 18 months) found that 38·7% had developed PU indicating imminent death; within this grouping 30% appeared at 0–3 months before death and 10% of the PUs appeared 9–12 months before death 24.

The natural course of advance dementia includes a prolonging of impaired mobility, accompanied with febrile episodes, infections, aspiration pneumonia, eating problem and tube feeding. These encourage the development of PU, which in turn promotes advanced dementia as a terminal stage 25.

Parkinson's disease and low consciousness conditions (expressed in the GCS as a low range, mean 11–15) are common neurodegenerative diseases of the elderly with increased prevalence occurring with the ageing process. As increased rigidity reduces the capability to move, the patient lies in a position striking heels or other parts of body on hard surfaces or the wheel chair. This is expressed in progressive deterioration through immobility, increased spasticity, swallowing difficulty and reduced cognition, all contributing significantly to PU development 26, 27.

Anaemia of chronic disease that is expressed as low Hb level was significantly more prevalent in the PU group. This type of anaemia represents an outcome of the combined influence of advanced systemic diseases including chronic renal failure, oncologic disease and systemic infection. Chauhan et al. concluded anaemia as a significant risk factor for PU development 28. However, Margolis et al. established anaemia as a risk factor in univariate analysis (RR: 1·7) but not in the multivariate analysis 19. The explanation for this discrepancy is that anaemia among the elderly in community dwellings is different from anaemia occurring in a skilled nursing department. Anaemia of chronic disease may express severity, progression and instability of the patient's medical condition that contributes significantly to PU development 29.

Requiring urinary catheters can be the result of PU severity as there is a caregiver tendency to insert urinary catheters to PU patients in the sacrum or the hip region in order to prevent further deterioration and to allow efficient treatment 30. However, the customary practice involving urinary catheters may also reflect the use of medical devices related to the development of PU 31.

Nutritional deficiencies including level of BMI and weight had a significantly lower proportion in this study in the PU group, indicating malnutrition, acuteness and severity of the patient's medical condition. In four studies, low weight was found as a risk factor for PUs 14, 16, 21, 22. Seven studies were identified as indicating nutrition as a risk factor for PU based on a systematic literature review 14, 16, 17, 18, 19, 20, 28.

Serum albumin and total protein reflect protein synthesis affected by other factors (not nutrition) including inflammation, hydration and wound severity resulting from the medical situation and the outcome of multiple advanced diseases of the patients 31. This study confirmed previous studies showing that serum albumin levels are lower in patients with PUs 32, 33.

A higher frequency of antibiotic treatment was employed significantly in the PU group. Thus, patients with fever and systemic infections (pneumonia, urinary, soft tissue and sepsis) deteriorate the clinical condition and aggravate the PU condition. Two studies indicate sepsis and fever as risk factors for PU 12, 18.

The lower length of stay in the unit and the low median survival of 94 days for the PU group patients indicated a lower period of remaining and often entailed by death. Thus, this study indicates that PU is associated with the existence of certain systemic factors including neurodegenerative conditions, anaemia, recurrent infection, low BMI, low weight and low albumin and also associated with geriatric conditions including functional instability and older age. These factors complicate and deteriorate the patient's clinical condition and aggravate the PU condition.

Therefore, the treatment of PU should be extended to the treatment of systemic factors (e.g. providing high protein nutrition, antibiotic for infection, blood for anaemia and angioplasty for peripheral vascular disease). Associated treatment of systemic disease in correlation with PU also focuses on hypothyroidism, oedema from congestive heart failure, metabolic disorder, agitation and geriatric conditions (e.g. relief spasticity, withdrawing urinary catheter and tube feeding in difficult swallowing.)

PUs constitute a geriatric syndrome that involves not only multisystemic factors but also local injury to the soft tissue. The term geriatric syndrome indicates multifactorial health conditions (advanced dementia, low BMI and Hb, existing systemic diseases or conditions), occurring when the cumulative effect of impairment in multiple organ systems renders the immobile and functionally impaired elderly patient vulnerable 6. The syndrome stresses multiple risk factors that synergistically converge to the appearance of PUs. The appearance of PUs is the final common pathway that accumulates and progresses over a course of years. The coexistent common risk factors act synergistically and even feed on each other, interacting with other geriatric conditions (e.g. frailty, delirium, incontinence and falls). Understanding the connection between PUs and the geriatric syndrome is the key to understanding why the frail and multicomorbidity elderly are vulnerable to PUs and equally permit targeted intervention for prevention and treatment.

Study limitations

Our study's limitations include the study's setting. Conducting a study in a skilled nursing unit rather than in the community or in nursing homes is obviously associated with a selection bias towards sicker patients and higher severity of PU (grades III and IV). Our control group is composed of patients who can apply for the department's admission criteria. Therefore, our results may not be generalisable to healthier ageing patients, and we cannot attribute the lower survival rate in the PU group solely to the presence of ulcers. Furthermore, we describe possible risk factors identified concomitantly with the diagnosis of PU or later during hospitalisation; therefore, the temporal association between these factors and the development of PU cannot be ascertained.

Conclusions and practice recommendations

Specific geriatric conditions and certain systemic diseases were found in high significant prevalence in the PU group. These multiple risk factors accumulate and interact synergistically in a concentric model during the patient's life and result in the final common pathway as the visible appearance of a PU syndrome.

Therefore, it is important to establish a comprehensive treatment (local, systemic, prevention and quality of life). The team is composed of multidisciplinary staff specially trained in geriatrics. The multidisciplinary staff is effective in recognising that patients with these systemic conditions are vulnerable to occurrence of PU and thus intervene with appropriate prevention strategies to delay appearance of PU. As PU is the potential accelerant for death, the goal of the treatment in these frail patients differs from healing of the wounds to concentrate on symptom control, prevention of complications and to contribute to their quality of life.