Modelling the cost‐effectiveness of subepidermal moisture measurement as part of a process of assessment and intervention to prevent hospital‐acquired pressure ulcers

Abstract

Skin tissue assessment is traditionally used to identify early signs of pressure damage from changes observed at the skin surface. However, the early onset of tissue damage induced by pressure and shear forces is likely to be on soft tissues beneath the surface of the skin. Subepidermal moisture (SEM) is a biophysical marker for the detection of early and deep pressure‐induced tissue damage. Measurement of SEM can detect early pressure ulcers up to 5 days before visible skin changes occur. The aim of this study was to evaluate the cost‐effectiveness of SEM measurement compared with visual skin assessment (VSA). A decision‐tree model was developed. Outcomes are the incidence of hospital‐acquired pressure ulcers, quality‐adjusted life‐years (QALYs) and costs to the UK National Health Service. Costs are at 2020/21 prices. The effects of parameter uncertainty are tested in univariate and probabilistic sensitivity analysis. In a representative NHS acute hospital, the incremental cost of SEM assessment as an adjunct to VSA is −£8.99 per admission, and SEM assessment is expected to reduce the incidence of hospital‐acquired pressure ulcers by 21.1%, reduce NHS costs and lead to a gain of 3.634 QALYs. The probability of cost‐effectiveness at a threshold of £30 000 per quality‐adjusted life year is 61.84%. Pathways that include SEM assessment make it possible to implement early and anatomy‐specific interventions which have the potential to improve the effectiveness of pressure ulcer prevention and reduce healthcare costs.

1. INTRODUCTION

UK clinical guidelines for the prevention and management of pressure ulcers in adults recommend regular skin assessment by a trained healthcare professional, repositioning, barrier creams and pressure redistributing devices for people judged to be at risk. ¹ The standard of care in skin tissue assessment (STA) involves a combination of risk assessment tools, visual assessment of the skin and clinical judgement. For adults who have been assessed as being at risk, the NICE guideline recommends repositioning at least every 6 h, and every 4 h for adults who are at high risk. ¹

Visual skin assessment relies on evidence of changes which can be observed at the skin surface. However, the early onset of tissue damage induced by pressure and shear forces is likely to be on soft tissues beneath the surface of the skin. Changes occurring at the cellular level manifest as damage to the subepidermal layers before skin changes become visible, and changes may not be detected until the damage has already occurred. The accuracy of assessments depends partly on the training and experience of the clinician, and the sensitivity and specificity of visual assessment in routine practice are relatively low (50.6% and 60.1% respectively). ² Visual assessment is particularly difficult in people with darker skin tones. ³ , ⁴

A localised change in subepidermal moisture (SEM), localised oedema or persistent focal oedema is a biomarker for the detection of an early pressure ulcer which precedes visible skin changes by an average of 5 days. ⁵ , ⁶ , ⁷ , ⁸ Pooled data from studies of residents in US nursing homes showed that measurement of subepidermal moisture identified erythema/category 1 pressure ulceration 1 week before it was detectable by visual assessment, including people with darker skin tones. ⁵ In an acute hospital setting, the mean time to detection of pressure damage by visual inspection was 5.5 days, compared with a mean of 1.5 days by measurement of subepidermal moisture. ⁷ A recent systematic review found than SEM measurements may detect early pressure damage up to 8 days earlier than visual assessment of the skin. ⁹ Early anatomy‐specific detection offers an opportunity to implement preventive interventions designed to stop or reverse the consequences of pressure damage.

When pressure damage develops, cell death leads to localised and persistent oedema which can be detected by measurement of subepidermal moisture. Comparison of SEM readings at different anatomical locations at and around the sacrum and heels are used to compute a delta value (Δ). ¹⁰ , ¹¹ A SEM delta ≥0.6 is indicative of early stages of pressure damage. ⁸

Skin tissue assessment and measurements of subepidermal moisture are tests which provide information to aid clinical decision‐making. The sensitivity of a test is the probability of a positive test result in people with pressure damage, and the specificity of a test is the probability of a negative test result in people without pressure damage. The aim is to maximise the ability of a test to identify the true positive cases, and at the same time minimise unnecessary costs by minimising the number of false positives. High sensitivity with high specificity results in a high proportion of true cases being detected with a relatively low number of false positives. ¹² The sensitivity and specificity of SEM assessment (at a SEM Δ ≥0.6) are 86.5% and 88.0%, respectively. ¹³

A test with superior predictive accuracy is not necessarily the most cost‐effective once consideration is given to the effect on patient outcomes and costs. A link must be established from test result to clinical decision and from clinical decision to patient and healthcare system outcomes. The aim of this study was to evaluate the cost‐effectiveness of subepidermal moisture assessment compared with standard of care by considering effects on the incidence of hospital‐acquired pressure ulcers, quality of life of patients and costs to the NHS.

2. MATERIALS AND METHODS

A decision tree model was developed in MS Excel to compare the annual costs and outcomes between different methods of skin assessment for a cohort of acute hospital admissions judged to be at risk of developing a pressure ulcer. The perspective of the analysis is costs to the NHS arising from pressure ulcers which are hospital acquired. The timescale is 1 year from hospital admission, and costs include inpatient costs of skin assessment, prevention, treatment of incident pressure ulcers and costs incurred in the community post‐discharge.

Subjects entering the decision tree (Figure 1) are hospital inpatients who have been assessed on admission by one of the standard risk assessment tools (such as Braden or Waterlow. ¹⁴ , ¹⁵ ) as being at risk of pressure ulceration. All are placed on a standard pressure‐relieving surface. Skin assessments are performed daily, and patients are repositioned every 6 h.

• Daily skin assessments result in one of two outcomes: a positive test leads to a decision to increase repositioning from four times to six times daily (every 4 h). A negative test leads to no change in the standard prevention protocol.

• A positive or negative test result may be true or false. False positive and true negative cases do not have pressure damage and incur no pressure ulcer treatment costs. True positives and false negatives do have pressure damage: some will resolve before a pressure ulcer develops, and others will require treatment.

• The probability that a person with underlying pressure damage will heal before discharge depends on whether the damage is detected by the test (true positive) and whether it is detected early (day 1–3) or late (day 3–7). Pressure damage detected first by measurement of subepidermal moisture will typically occur between 4 and 6 days before changes are visible on the surface of the skin.

• Pressure ulcers which are not resolved incur the costs of pressure ulcer treatment, depending on the category of the ulcer.

FIGURE 1.

Decision tree.

Standard of care in skin tissue assessment (STA) includes risk assessment tools, visual assessment of the skin and clinical judgement. The intervention is measurement of subepidermal moisture (SEM) with a cut‐off delta value ≥0.6 in place of, or as an adjunct to, visual skin assessment (VSA). Both the intervention and control include prior risk assessment and clinical judgement. Cost‐effectiveness is evaluated for two possible decision rules:

• VSA vs SEM. Clinical decisions about enhanced interventions to prevent ulceration are based on visual skin assessments alone or on measurements of subepidermal moisture alone.

• VSA vs dual assessment (VSA plus SEM). When SEM is an adjunct to visual assessment, decisions are made on visual skin assessment alone or on a combination of skin tissue assessment and measurement of subepidermal moisture. With dual assessment, the decision rule will classify a positive test result (tissue damage present) if either VSA or SEM registers a positive result.

Cost‐effectiveness is assessed by comparing the incremental cost per quality‐adjusted life year (QALY) for the intervention (incremental cost‐effectiveness ratio ICER) compared with VSA alone against a threshold of £30 000 per QALY gained. In the UK the National Institute for Health and Care Excellence (NICE) uses a cost‐effectiveness threshold range between £20 000 and £30 000 per QALY gained. ¹⁶ The threshold is a proxy for the value of health gains forgone because when budgets are fixed, a decision to adopt a new technology implies a reduction in spending on other treatments. Outcomes are also expressed as net monetary benefit (NMB) which values QALY gains at £30 000 per QALY. Because of the annual timescale of the analysis, costs and outcomes are not discounted.

2.1. Sensitivity analysis

To reflect local variation between hospitals, or between wards, pressure ulcer incidence rate, average length of inpatient stay, frequency and cost of skin assessments, and pressure ulcer treatment costs are varied in a univariate sensitivity analysis.

Parameter uncertainty is also explored in a probabilistic sensitivity analysis (PSA). Distributions were assigned to each of the main parameters: Beta distribution for parameters with a range of 0.0–1.0 (test sensitivity and specificity, incidence of hospital acquired pressure ulcers, and rates of ulcer healing); Gamma distribution for other variables (costs, nurse time and health‐related quality of life). A conservative 25% standard error was assumed where the standard error was unknown. Distributions were sampled at random, and the results of the cost‐effectiveness analysis were recorded for each of 10 000 Monte Carlo simulations.

Outcomes of the PSA are the probability that an intervention is cost‐effective at a threshold of £30 000 per QALY gained, the probability an intervention is cost saving, and the probability an intervention leads to an improvement in the quality of life of patients.

3. DATA SOURCES

Parameter values are summarised in Table 1. The model is calibrated for a representative NHS acute hospital with 450 beds, 83% average occupancy rate and an average length of stay 4.5 days. ²⁵ , ²⁶ 41% of 30 295 potential annual admissions are assumed to be at risk of pressure ulceration, ¹⁷ giving a cohort of 12 421 annual admissions entering the model.

TABLE 1.

Model parameters.

Parameter	Parameter value	Source
Cohort entering the model	12 421 patients at risk of pressure ulceration (Beds = 450; occupancy rate = 83%; average length of stay = 4.5 days; proportion at risk = 41%)	NHS England, SDCS data collection NHS Digital, Hospital Episode Statistics Vanderwee 17
Incidence of hospital‐acquired pressure ulcers	3.58%	Clark 18 ; Smith 19
Visual skin assessment (VSA)	Sensitivity = 50.6% Specificity = 60.1%	Pancorbo‐Hidalgo 2
SEM scanner	Sensitivity = 86.5% Specificity = 88.0%	Gershon 13
Joint VSA plus SEM	Sensitivity = 93.3% Specificity = 52.9%	Calculated by combining VSA and SEM assuming conditional independence
Single use SEM scanner per patient per day	£5.50	Manufacturer
Nurse cost	£41/h Band 5 hospital nurse	PSSRU 20
Standard prevention cost per day	£54.67	Assumption. 2 Band 5 nurses 10 min each, every 6 h
Enhanced prevention cost per day	£82.00	Assumption. 2 Band 5 nurses 10 min each, every 4 h
Nurse time per VSA	£3.42	Assumption. 1 Band 5 nurse 5 min
Nurse time per scan	£3.42	Assumption. 1 Band 5 nurse 5 min
Category distribution of incident pressure ulcers	Category I: 32.9% Category 2: 49.5% Category 3: 12.2% Category 4: 5.4%	Clark 18
Treatment costs by category of pressure ulcer	Category I: £1536 Category 2: £9627 Category 3: £10 795 Category 4: £11 184	Guest 21 at 2020/21 prices (PSSRU NHS cost inflation index (NHSCII)
Healing rate of detected category 1 pressure ulcer	53.7%	Halfens 22
Healing rate of undetected category 1 pressure ulcer	32.26%	Halfens 22
Utility decrement (ED‐5D)	Category 1: 0.009 Category 2: 0.052 Category 3: 0.087 Category 4: 0.106	Palfreyman 23 ; Bennett 24

The incidence of hospital‐acquired pressure ulcers including category 1 is not routinely recorded in the UK. An estimate was derived by pooling point prevalence rates from an audit of 24 acute hospitals in England and a survey of 66 hospitals in Wales covering 10 604 patients (8.53% including category 1 pressure ulcers) ¹⁸ , ¹⁹ combined with an estimate of the proportion of prevalent pressure ulcers which were hospital‐acquired (41.95%), ¹⁹ to give a baseline incidence estimate of 3.58%. These are true positive cases which have been identified by visual assessment. The category distribution of incident ulcers is an estimate based on a national study of acute and community hospital patients in Wales designed to identify the prevalence of pressure ulcers. ¹⁸ The survey was carried out in 1 month in 2015 and covered 8365 patients in 66 hospitals. Classified by the most severe pressure ulcer: 32.9% were category 1; 49.5% category 2; 12.2% category 3; and 5.4% category 4.

Costs of prevention include only the costs of nurse time required to reposition the patient, assuming that specialist pressure‐relieving equipment is available in the hospital, so the analysis does not include the opportunity costs of allocating equipment to one patient rather than another when equipment is scarce. Standard prevention protocols are assumed to require two Band 5 nurses for 10 min each, four times daily (£54.67 daily). The cost of enhanced (four‐hourly) repositioning is £82.00 daily. The cost of a Band 5 hospital nurse including overheads is £41/h. ²⁰

The Provizio® SEM scanner (Bruin Biometrics, LLC) is a hand‐held device that measures subepidermal moisture at and around the bony prominences on the heels and sacrum. The scanner compares minimum and maximum readings at each site to calculate the difference (SEM delta‐Δ). A delta value ≥0.6 is indicative of the presence of underlying tissue damage. The cost of SEM assessment is £5.50 per patient per day for a single‐use sensor which is attached to a hand‐held device provided free of charge by the manufacturer. The nurse time cost of visual assessment and scanning is the same, assuming one Band 5 hospital nurse for 5 min at £3.42 per assessment. Skin assessments are assumed to be carried out once daily for the duration of the inpatient admission.

The sensitivity (50.6%) and specificity (60.1%) of nurses' clinical judgement are derived from a systematic literature review of risk assessment scales for pressure ulcer prevention. ² The relatively low sensitivity may be due in part to the difficulty of distinguishing between a category 1 pressure ulcer and redness or other marks on the surface of the skin. The sensitivity and specificity of spatial variation in SEM readings were evaluated in a population of residents in residential or nursing homes with confirmed category 1 or 2 pressure ulcers (Arm 1, n = 125), and a second population without an existing pressure ulcer or broken skin (Arm 2, n = 50). Readings were taken at different locations at and around the sacrum and heels. Aggregate (sacrum and heel) sensitivity and specificity at a SEM delta ≥0.6 were 86.8% (112/129) and 88.0% (88/100). ¹³ The specificity of SEM readings may be underestimated because some of what appear to be false positives are true positive cases in which the underlying pressure damage resolves before it becomes visible.

Assuming tests are conditionally independent, the sensitivity and specificity of a combination of tests, in this case VSA plus SEM, can be computed using the multiplication rule for joint probabilities. ²⁷ The decision rule in this case would classify a test result as positive if either VSA or SEM indicates the presence of underlying tissue damage. The sensitivity of this type of test will be higher than the sensitivity of either test taken on its own. But the test would also be expected to indicate a larger number of false positives, and hence lower specificity. The computed sensitivity and specificity of a joint test are 93.3% and 52.9%, respectively.

In a prospective study observing the clinical course of category 1 pressure ulcers in patients in acute care hospitals, the proportion of category 1 ulcers detected early which either resolved or healed within 3 days was between 53.7% and 63.5%. ²² The proportion of ulcers unhealed after 3 days which resolved or healed by day 7 was between 18.99% and 32.26%. The model assumes a healing rate of 53.7% for potential category 1 pressure ulcers detected early and 32.26% for pressure ulcers detected 3–4 days later.

Annual NHS pressure ulcer treatment costs by category of ulcer were derived from a retrospective cohort analysis of 209 patients with a pressure ulcer identified within a randomly selected population of 6000 obtained from the Health Improvement Network (THIN) database. ²¹ The analysis recorded details of (mainly) community healthcare resource use and costs from initial presentation for 12 months. These costs are a reasonable reflection of costs to the NHS because a majority of pressure ulcers acquired during an inpatient episode are likely to be managed in the community post‐discharge. Costs in the original source are at 2015/16 prices. Costs in the model are adjusted to 2020/21 prices using the NHS cost inflation index ²⁰ : category 1 £1536; category 2 £9627; category 3 £10 795; and category 4 £11 184 per episode.

The quality‐of‐life impact of pressure ulcers (Table 2) was measured in a sample of UK patients in acute and community‐based healthcare settings who either had a pressure ulcer or were at risk. The Euroqol instrument (EQ‐5D‐3L) was administered in person to patients in the acute sector (n = 273) and by postal survey for patients in community care (n = 34). ²³ A utility decrement was calculated by subtracting the mean EQ‐5D score for patients with different categories of pressure ulcers from the mean utility score of patients without pressure ulceration. An average QALY decrement by category of ulcer was calculated by applying estimates of pressure ulcer duration from Bennett ²⁴ (Table 2).

TABLE 2.

Health state utility by category of pressure ulcer.

Pressure Ulcer category	EQ‐5D mean score 23	Mean duration (days) 24	QALY decrement
No pressure ulcer	0.364
Category 1	0.250	28.4	0.009
Category 2	0.160	93.8	0.052
Category 3	0.115	127.4	0.087
Category 4	0.115	154.7	0.106

4. RESULTS

Results are for a cohort of 12 421 annual admissions at risk of pressure ulceration. The number of pressure ulcers in this population which would develop without intervention (n = 879) depends on the underlying incidence rate and is the same irrespective of the choice of assessment method. The differences are in the number of positive test results and the accuracy of a test in correctly identifying subjects with underlying pressure damage. False positives are a key driver of costs because a false positive assessment leads to enhanced preventive measures which are unnecessary. In the VSA scenario a false positive is when a patient is wrongly flagged by visual assessment as having a pressure ulcer when no pressure ulcer exists (for example, where it is difficult to distinguish between redness of the skin and a category 1 ulcer). In the SEM scenario a false positive is when SEM indicates the presence of pressure damage when no pressure damage is present.

4.1. VSA vs SEM

A decision rule based on visual skin assessments has a sensitivity of 50.6%, meaning that around half of the true cases are identified by the test (445/879). A specificity of 60.1% means 6937 of 11 542 true negative cases are identified. The total number of positive test results is 5050 of which 4605 (91%) are false positives (Table 3). A decision rule based on a SEM delta ≥0.6 would result in a smaller number of positive test results (2145) and would identify a larger proportion of true positive cases (760/879). The proportion of false positives is 64.6% (1385/2145) (Table 3).

TABLE 3.

Population disposition according to the presence of pressure damage and test result (VSA vs SEM).

		TEST
VSA		Positive	Negative
TRUE	Positive	445	434	879
	Negative	4605	6937	11 542
	ALL	5050	7371	12 421
SEM		TEST
		Positive	Negative
TRUE	Positive	760	119	879
	Negative	1385	10 157	11 542
	ALL	2145	10 276	12 421

The larger number of subjects testing positive with VSA (including false positives) compared with SEM leads to more expenditure on prevention (£3.677 m vs £3.319 m) and the lower proportion of true positive cases detected leads to fewer ulcers prevented and higher treatment costs (£3.596 m vs £3.109 m). The additional cost of SEM devices, training and scanning is £0.509 m compared with a cost for nurse VSA assessments of £0.191 m.

The incremental cost of SEM compared with VSA is −£42.29 per person (7.0%); −£525 244 for the cohort (Table 4). The number of pressure ulcers prevented in the SEM scenario is higher by 68 (17.9%), which leads to a gain of 3.055 QALYs. SEM is expected to be a dominant option compared with VSA. The PSA shows that the SEM option has 81.88% probability of being cost‐effective relative to VSA at a threshold of £30 k/QALY, a high probability of being cost saving (79.68%) and a high probability of generating more QALYs (90.46%) (Table 4). Figure 2 shows the cost‐effectiveness plane for a threshold of £30 000 per QALY gained.

TABLE 4.

Cost‐effectiveness of VSA vs SEM (2020/21 prices).

	VSA	SEM	SEM‐VSA
Cost per patient	£601	£559	−£42.29
Total cohort cost	£7 463 191	£6 937 947	−£525 244
Total QALYs	4498.64	4501.70	+3.055
ICER			SEM dominant
Pressure ulcers prevented	378.8	446.5	68 (17.9%)
Probability of cost‐effectiveness			0.8188
Probability of cost saving			0.7968
Probability of higher QALYs			0.9046
Expected net monetary benefit			£726 436

FIGURE 2.

Cost‐effectiveness plane, VSA alone vs SEM.

4.2. VSA vs dual assessment

A decision rule based on a dual assessment has a sensitivity of 93.3%, which means that a very high proportion of positive cases would be identified, and a dual assessment will produce a high number of positive results (n = 6256). However, because the specificity of dual assessment is low (52.9%) a large proportion of those positive test results (86.9%) will be false (Table 5). The result is higher prevention costs than VSA alone (£3.825 m vs £3.677 m), but lower costs of pressure ulcer treatment (£3.017 m vs £3.596 m).

TABLE 5.

Population disposition according to the presence of pressure damage and test result (dual assessment).

		TEST
Dual assessment		Positive	Negative
TRUE	Positive	820	59	879
	Negative	5436	6106	11 542
	ALL	6256	6165	12 421

The incremental cost of dual assessment compared with VSA alone is −£8.99 per person (1.5%); −£111 725 for the cohort (Table 6). The number of pressure ulcers prevented in the dual assessment scenario is higher by 80 (+21.1%), leading to a gain of 3.634 QALYs. The probability of cost‐effectiveness is 61.81% and the probability the dual combination is costsaving is 57.0%. The probability of higher QALYs is 90.7%. The addition of SEM to standard practice is expected to lead to a small cost saving and better outcomes. Figure 3 shows the cost‐effectiveness plane for a threshold of £30 000 per QALY gained.

TABLE 6.

Cost‐effectiveness of VSA vs dual assessment (VSA + SEM) (2020/21 prices).

	VSA	Dual assessment	Dual assessment‐
Cost per patient	£601	£592	−£8.99
Total cohort cost	£7 463 191	£7 351 466	−£111 725
Total QALYs	4498.64	4502.27	+3.634
ICER			Dual assessment dominant
Pressure ulcers prevented	378.8	459.3	80 (21.1%)
Probability of cost‐effectiveness			0.6184
Probability of cost saving			0.5700
Probability of higher QALYs			0.9140
Expected net monetary benefit			£365 968

FIGURE 3.

Cost‐effectiveness plane, VSA alone vs VSA + SEM.

Results suggest that both SEM options are cost‐effective compared with VSA alone, but a comparison of SEM alone with dual assessment suggests that a decision rule based on a joint test is unlikely to be cost‐effective compared with a decision based on SEM alone. The joint test prevents 12 more ulcers (80–68) and generates an additional 0.579 QALYs. However, the incremental cost of dual assessment compared with SEM alone is £413 519, giving an ICER greater than £700 000 per QALY, well outside the usual willingness‐to‐pay threshold in the UK.

5. UNIVARIATE SENSITIVITY ANALYSIS

To reflect local variation between hospitals, or between wards, frequency of repositioning and/or the number and grade of staff required; average length of stay; costs of pressure ulcer treatment; and pressure ulcer incidence were varied in a univariate sensitivity analysis (Table 7).

TABLE 7.

Univariate sensitivity analysis (2020/21 prices).

		VSA vs SEM		VSA vs dual assessment
	SA value	Pressure ulcers prevented. QALYs gained	Incremental cost pp	Pressure ulcers prevented. QALYs gained	Incremental cost pp
Base case		68 3.055	−£42.29	80 3.630	−£8.99
Repositioning £246 pp (6 h) £369 pp (4 h)	+100% £492 £738	No change	−£71.05	No change	+£2.98
Length of stay 4.5 days	+100% 9 days	No change a	−£46.30	No change a	+£27.7
Treatment costs £1536 £9627 £10 795 £11 184	+50% £2304 £14 440 £16 192 £16 776	No change	−£61.87	No change	−£32.29
Incidence 3.58%
General acute	5.6%	106 4.779	−£61.25	126 5.685	−£33.54
ICU	17%	321 14.509	−£168.28	382 17.257	−£172.03
Spinal injury Long‐term elderly	30%	567 25.604	−£290.33	674 30.453	−£329.96

^a Changing length of stay alone is misleading because it does not take account of changes in pressure ulcer risk which may accompany longer stays.

In the base case analysis one of the benefits of the SEM option as a substitute for VSA is a reduction in the number of patients testing positive, and this has the effect of reducing the number requiring enhanced prevention. Higher unit costs of prevention activities increase the cost savings associated with the SEM option. For example, a doubling of repositioning costs per episode reduces the incremental cost of SEM from −£42.29 pp to −£71.05 pp (Table 7). When SEM is an adjunct to standard care, in the base case the number of test positive cases is higher than with VSA alone. In this case doubling of prevention costs per patient leads to a net incremental cost of £2.98 per patient (Table 7). The breakeven value of prevention costs is 75% above the base case value. Below this level the dual assessment option is still cost saving compared with VSA.

So long as an intervention reduces the number of pressure ulcers requiring treatment, higher per ulcer treatment costs lead to a higher cost saving with SEM irrespective of whether SEM is a substitute for VSA or an adjunct, because fewer ulcers are treated. For example, increasing treatment costs for each grade of pressure ulcer by 50% leads to savings in costs per person with the SEM/dual assessment option of £61.87 (VSA vs SEM) or £32.29 (VSA vs dual assessment). The level of treatment costs at which options become cost neutral is at 10% of the base case value (VSA vs SEM) and less than 0.8% (VSA vs dual assessment).

Differences in the average length of stay impact on costs by increasing the length of time a patient needs to be repositioned and by increasing the daily costs of SEM assessments. Incremental costs are marginally lower with longer length of stay for VSA vs SEM (−£46.30) and higher for the dual assessment option (Table 7). For example, doubling the average length of stay results in a net cost of −£46.30 (VSA vs SEM) and + £27.70 (VSA vs dual assessment). However, the model does not take account of the fact that a longer length of stay increases the length of time a patient is at risk, and settings such as residential or nursing home care with longer length of stay may also be associated with higher incidence.

Pressure ulcer incidence is an important driver of results which is likely to vary between and within hospitals, and between different care settings. For a constant population at risk, higher incidence is associated with greater cost savings with SEM because more ulcers are prevented, and the effect is substantial. SEM is less cost‐effective in populations with very low risk (<1%) (Table 7).

The assumed incidence rate in the base case is conservative. Estimates of pressure ulcer incidence aggregated across all the specialties in an acute hospital vary. A systematic review and meta‐analysis of multinational studies reporting the prevalence and incidence of pressure injuries in hospitalised patients identified 42 relevant studies covering more than two million patient episodes. The pooled rate of hospital‐acquired pressure injuries was 8.4% (95% CI 7.6%‐9.3%). Category 1 pressure ulcers accounted for 43.5% of the total. ²⁸ Two studies reporting incidence in UK acute hospitals recorded rates of 2.1% and 1.7% for ≥ category 2 pressure ulcers. ²⁹ , ³⁰ Adjusting to include category 1 (at 40%) gives overall estimates of pressure ulcer incidence 3.5% and 2.8%, respectively. Assuming a mid‐point estimate of incidence of 5.6% for a general acute hospital population, the number of pressure ulcers prevented would be 106 and 126, and the incremental cost of the intervention would be −£61.25 pp and −£33.54 pp for VSA vs SEM and VSA vs dual assessment, respectively (Table 7).

Certain patient populations are likely to have higher than average risk. People with darker skin tones are more likely to develop pressure injuries which are difficult to detect by visual assessment alone. ³¹ , ³² In a population of 1938 residents aged 65 or older newly admitted to one of 59 nursing homes in Maryland, Baumgarten ³³ reports a pressure ulcer incidence rate 1.31 times higher in patients with dark skin tone compared with other patients. Applying this ratio to the model, the baseline incidence for patients with darker skin tone would be 4.69%. SEM assessment is expected to lead to between 89 and 105 more pressure ulcers prevented with SEM/dual assessment compared with VSA alone, and cost savings between £52.71 and £22.48 per patient.

Patients in intensive care are particularly susceptible to pressure damage because of immobility and the presence of significant comorbidity. A systematic review and meta‐analysis of incidence and prevalence studies in adult intensive care identified 10 studies reporting the incidence of pressure injuries. The 95% confidence interval for estimates of cumulative incidence was 10.0% to 25.9%. ³⁴ An international point prevalence study with follow‐up until hospital discharge recorded data from 13 254 patients in 1117 ICUs in 90 different countries. ³⁵ ICU‐acquired pressure ulcer incidence was 16.2% (95% CI 15.6%‐16.8%) of which 11.0% (95% CI 10.5%‐11.5%) were ≥ category 2. At a representative estimate for ICU units of 17%, the number of pressure ulcers avoided would be 321 and 382, and the incremental cost of the intervention would be −£168.28 pp and −£172.03 pp for VSA vs SEM and VSA vs dual assessment, respectively (Table 7).

Residents in nursing homes or long‐term residential care and spinal injury patients are also populations at high risk. Over a 28‐day period the incidence of non‐blanchable erythema in a population of 838 nursing home residents was 34.8% to 38.1%, depending on the turning regime. The incidence of category 2–4 ulcers was between 3% and 24.1%. ³⁶ A two‐arm randomised controlled trial in 16 Belgian elderly nursing homes (n = 235) was designed to assess the effectiveness of different turning regimes on pressure ulcer incidence. The incidence of category 2–4 pressure ulcers was 16.4% and 21.2% in the experimental and control groups, respectively. ¹⁷ Adjusting to include category 1 (at 40%), the overall incidence rates would be 27.3% and 35.3%. Pooled results from a meta‐analysis of 24 studies including 600 000 patients with spinal injury found an incidence rate of 32.36% (95% CI 28.21%‐36.51%). ³⁷ Assuming an incidence rate in these populations of 30%, the number of pressure ulcers avoided would be 567 and 674, and the incremental cost of the intervention would be −£290.33 and −£329.96 for VSA vs SEM and VSA vs dual assessment, respectively.

6. DISCUSSION

Many pressure ulcers are preventable, but effective prevention relies on early detection. Regular visual assessment of the skin has been the traditional method of skin assessment, but pressure damage can be present before it becomes visible. Measurement of subepidermal moisture is a biomarker for the development of a pressure ulcer which precedes visible skin changes by an average of five days. Information from SEM measurements should make it possible to implement enhanced preventive measures earlier, leading to a reduction in the incidence of pressure ulcers. The aim of this analysis was to evaluate the cost‐effectiveness of implementing SEM assessment into routine care pathways using a simple decision tree model. Results suggest that SEM assessment could be used to improve patient outcomes, reduce the costs of unnecessary preventive interventions, and reduce the costs of pressure ulcer treatment.

Compared with visual assessment alone, SEM assessment is expected to result in a 17.9% reduction in the incidence of pressure ulcers and a cost saving of £42.29 per person. Results are robust to variation in parameter values: the probability that SEM is cost‐effective compared with VSA alone at a threshold of £30 000/QALY is 81.88%.

In a comparison of visual assessment and dual assessment, the joint decision rule leads to a larger reduction in the incidence of pressure ulcers (21.1%) but at a higher cost. The cost saving per person in this scenario is £8.99, and the probability of cost‐effectiveness is 61.84%. The joint decision rule leads to more positive test results which lead to more cases being prevented, but also to a higher number of false positives.

Two previous studies have evaluated the cost‐effectiveness of a SEM scanner as an adjunct to standard practice. A decision‐tree model similar in structure to the present model evaluated the cost‐effectiveness of a SEM scanner as an adjunct to visual skin assessment, compared with visual skin assessment alone from a UK acute care perspective. ¹¹ Assuming an inpatient incidence rate of 1.6%, the model predicts a cost saving with the scanner of £15.23 per admission, and £80.88 assuming an incidence rate of 6.3%. ¹¹ In another study, a Markov simulation model was developed to assess the cost‐effectiveness of SEM compared with standard prevention guidelines from a US perspective. ³⁸ The integration of SEM assessment into clinical practice was estimated to lead to a reduction of US$4054 per patient and a gain of 0.35 quality‐adjusted life years. For every 1000 admissions of high‐risk patients, introduction of scanning could avert around seven hospital‐acquired pressure‐injury associated deaths and save around 206 bed‐days.

The results of a modelling approach depend on assumptions about how information from a test changes clinical decisions, and on how changes in clinical decisions translate into a reduction in the incidence of pressure ulcers. This is the main limitation of a modelling approach compared with a comparative study carried out in normal clinical practice. However, the conclusions of our approach are consistent with real‐world evidence reported in the literature.

Our modelling predicts a reduction in category 1 to 4 pressure ulcer incidence of between 17.9% and 21.1%. Four clinical studies report a reduction in incidence in clinical practice. A pre‐ and post‐intervention study enrolled patients from 28 institutions in the UK, Belgium, Canada, Spain and Ireland with the aim of evaluating the effectiveness of subepidermal scanning as an adjunct to standard practice in reducing the incidence of category 2 to 4 pressure ulcers. ³⁰ Pooling data from all the sites, the overall relative risk (RR) of ulceration in the intervention period was 0.38 (95% CI: 0.26‐0.56), which represents a reduction of 62% compared with the pre‐intervention period. A comparison of incidence rates before and during the introduction of the SEM scanner was carried out as part of a routine audit in 15 centres in the UK: 13 of which were acute hospitals. ²⁹ Patients had a daily visual assessment and SEM scanning on the sacrum and heels. A SEM Δ ≥ 0.6 alerted nurses to an increased risk of pressure ulceration and additional early interventions were provided. All the study sites reported a reduction in incidence over the intervention period compared with the pre‐intervention period. The overall reduction in category 2 to 4 incidence was 87.2% (acute), 46.7% (palliative care), and 26.7% (community care). Patients with a Waterlow score ≥ 10 (n = 697) were enrolled across four wards in an acute hospital in the UK over a 6‐month period. The aim was to compare the incidence of category 2 to 4 pressure ulcers before and after the addition of subepidermal scanning to standard clinical practice. In the 12 months prior to the intervention, the incidence rate of category 2 to 4 pressure ulcers was 1.48%, in the study period the rate was 0.287%, a reduction of 80.6%. ³⁹ Raizman ⁴⁰ reports the results of a two‐part observational study. In phase 1, patients received standard of care risk assessments and interventions and were scanned with the SEM scanner. Clinicians were blinded to the SEM reading. In part 2, patients received the same assessments, but clinicians were able to use information from the SEM scan to inform judgements about appropriate interventions. In phase 1, 12/89 subjects (13.5%) developed a category 1 to 4 pressure ulcer. In phase 2, 2/195 subjects (1%) developed an ulcer, which represents a reduction of 92.5%.

The implications for clinical practice are important. Introducing measurement of subepidermal moisture as an adjunct to visual skin assessment is expected to be a cost‐effective addition to current practice. Replacing visual assessment with measurement of subepidermal moisture is the most cost‐effective option, but a decision rule which is based on information from both visual assessment and SEM provides more benefits for patients but at a higher cost.

CONFLICT OF INTEREST STATEMENT

The authors declare that there is no conflict of interest.

Posnett JW, Moss JWE, Michaelwaite LI. Modelling the cost‐effectiveness of subepidermal moisture measurement as part of a process of assessment and intervention to prevent hospital‐acquired pressure ulcers. Int Wound J. 2023;20(7):2688‐2699. doi: 10.1111/iwj.14143

DATA AVAILABILITY STATEMENT

Data obtained from publicly available sources: published literature and national NHS datasets.