p u b l i c h e a l t h 1 2 8 ( 2 0 1 4 ) 7 7 1 e7 7 6

Available online at www.sciencedirect.com

Public Health journal homepage: www.elsevier.com/puhe

Original Research

The carbon footprint of acute care: how energy intensive is critical care? A.S. Pollard a,b,c, J.J. Paddle b, T.J. Taylor a,*, A. Tillyard b a European Centre for Environment and Human Health, University of Exeter Medical School, Knowledge Spa, RCH Treliske, Truro TR1 3HD, UK b Royal Cornwall Hospitals NHS Trust, Truro TR1 3LJ, Cornwall, UK c Pollard Systems Ltd, Mevagissey, Cornwall PL26 6TL, UK

article info

abstract

Article history:

Objectives: Climate change has the potential to threaten human health and the environ-

Received 4 July 2013

ment. Managers in healthcare systems face significant challenges to balance carbon

Received in revised form

mitigation targets with operational decisions about patient care. Critical care units are

14 May 2014

major users of energy and hence more evidence is needed on their carbon footprint.

Accepted 12 June 2014

Study design: The authors explore a methodology which estimates electricity use and

Available online 2 September 2014

associated carbon emissions within a Critical Care Unit (CCU). Methods: A bottom-up model was developed and calibrated which predicted the electricity

Keywords:

consumed and carbon emissions within a CCU based on the type of patients treated and

Carbon footprint

working practices in a case study in Cornwall, UK.

Critical care

Results: The model developed was able to predict the electricity consumed within CCU with

Health planning

an error of 1% when measured against actual meter readings. Just under half the electricity

Numerical analysis (computer

within CCU was used for delivering care to patients and monitoring their condition.

assisted)

Conclusions: A model was developed which accurately predicted the electricity consumed

Health informatics

within a CCU based on patient types, medical devices used and working practice. The model could be adapted to enable it to be used within hospitals as part of their planning to meet carbon reduction targets. © 2014 The Royal Society for Public Health. Published by Elsevier Ltd. All rights reserved.

Introduction Climate change has been identified by some as the greatest threat to human health in the 21st century.1 However, efforts to mitigate emissions of greenhouse gas (GHG) have been slow

to emerge.2 In the USA healthcare comprises 7% of total GHG emissions.3 In England, this fraction is more than 3%,4 reaching a total of 21 million tonnes of CO2 equivalent in 2010.5 The National Health Service (NHS) has a self-imposed target to lower GHG emissions by 80% by 2050.4 It is often considered within the NHS that mitigation against climate

* Corresponding author. European Centre for Environment and Human Health, University of Exeter Medical School, Knowledge Spa, RCH Treliske, Truro TR1 3HD, Cornwall, UK. Tel.: þ44 (0) 1872 258146; fax: þ44 (0) 1872 258134. E-mail address: [email protected] (T.J. Taylor). http://dx.doi.org/10.1016/j.puhe.2014.06.015 0033-3506/© 2014 The Royal Society for Public Health. Published by Elsevier Ltd. All rights reserved.

772

p u b l i c h e a l t h 1 2 8 ( 2 0 1 4 ) 7 7 1 e7 7 6

change is achievable in part through a reduction of the carbon footprint of hospitals.6 Previous studies have evaluated GHG emissions associated with secondary healthcare7 and renal care.8 The Critical Care Unit (CCU) in acute hospitals is a major consumer of energy, and as such it is important to consider options for improving efficiency in acute care.9 The objective of this study was, therefore, to build a model to estimate the electricity consumption and the associated carbon footprint in a CCU based on the patient caseload. The model may then be employed to quantify the energy savings achieved through different ways of delivering care. The suitability of this approach was then assessed as a tool to minimize the carbon footprint of CCUs.

Methods A bottom up model was developed based on observation of use of the CCU at the Royal Cornwall Hospital in the UK. This is a district general hospital of 750 in-patient beds, serving a population of approximately 450,000.10,11 The CCU has 10 funded adult beds, with a mixed medical and surgical casemix. Data were collected using chart review in terms of organs supported for all patients on the CCU from 14th May to 11th June 2012, during which time the electricity supplied to the unit was also measured. Knowing the rate at which each machine consumed electricity, the electricity consumed by estimating the proportion of time that machines were being used for each patient was modelled. The electricity used is supplied from two independent systems; the first supplies medical and bedside devices and the second supplies resources shared within the Unit. To reflect this, the electricity consumed was categorized within the Unit as:  Bedside Energy, Eb e that required to power devices for organ support and for patient monitoring;  Remote Energy, Er e that used by equipment that is not allocated on medical grounds to an individual patient (for example: blood gas analysis at the point of care), for lighting in corridors and other communal areas, and for information technology. The model estimates the energy required for ‘bedside’ purposes by analysing pathways of care. The ‘remote’ energy may be derived from consideration of the electricity required to power the shared space of the CCU, the thoroughfares, shared services and staff areas, which when added to the ‘bedside energy’, leads to an estimation of the Total Energy, E, expended in the Unit: E ¼ Eb þ Er

The model To quantify the bedside energy required, the model assesses energy demands for specific devices used in care pathways, which are labelled according to the illness severity and the treatment administered. The organs supported are:

    

Heart; Lungs; Gut; Kidneys; and Brain.

Four levels of support are available on the Unit which depict severity:12  Level 3 e Patients requiring advanced respiratory support, or basic respiratory support together with support of at least two organ systems. Generally these will be patients with complex conditions requiring treatment that can only be delivered by a CCU;  Level 2 e Patients requiring detailed observations or interventions including support for a single failing organ. This also includes patients requiring postoperative care and those ‘stepping-down’ from level 3 care. These patients also require care by a CCU;  Level 1 e Patients at risk of deterioration or those recently relocated from a higher level of care. Whilst the care of these patients can be met on an acute ward, they often require support from the critical care team; and  Level 0 e These patients can be managed on a normal ward. These care pathways are sufficiently distinguishable to allow the assignment of a unique combination of devices to each. A patient with complex needs may receive more than one type of support. In this case the patient would receive the care pathway appropriate to each organ simultaneously. The model automatically receives patient volumes through links established with sources of patient records. Data relating to expected machine usage for each patient type treated during a given period is entered into lists which are automatically fed into the model. This enables a calculation of the total bedside energy (Eb) expended for each patient according to the formula: Eb ¼

5 X 3 X i¼1

ri;j pi;j ti;j

j¼1

Where i: is the organ supported j: is the level of organ support ri,j: is the rating of the machine used in supporting organ i at level j in kW ti,j: is the total time that the patient received level j support to organ i pi,j: is the percentage of ti,j that the machine with rating ri,j was in use For example, the equipment used in treating a patient undergoing Level 3 support for lung and kidney for two days and the associated energy use is shown in Table 1. This calculated Eb is then summed across all patients treated in the Unit and compared against actual readings. Capacity-based ‘remote’ electricity usage, Er, is then added to complete the energy consumption within the Unit:

773

p u b l i c h e a l t h 1 2 8 ( 2 0 1 4 ) 7 7 1 e7 7 6

Table 1 e The Total Bedside Energy (Eb) expended in treating a typical patient in CCU across two days. This example contains a patient receiving Level 3 support for lungs and kidney. Machine used

Rating (W)

Pi,j

Consumption (kWh)

Ventilator Renal machine Humidifier Touch Screen Operator Monitor

125 200 220 20

100% 100% 100% 100%

6.00 9.60 10.60 0.96

44

100%

2.11

Total

609

Electricity consumed (kWh) in treating patient over 2 days

29.20

E ¼ Eb þ Er Each item, which when used and contributes to Er, has a rating allocated to it showing the electricity it consumes per unit time when it is running. Its duration of use can be estimated through consideration of staff working practices, drawing on the knowledge of senior staff within the Unit. Most equipment was running at all times, with the exception of lighting in bed bays (used 66% of the time) and computing equipment and lighting in administrative areas (used 33% of the time). Some of the items which contribute to E in the Unit are listed in Table 2.

Results Across the study period, a total of 38 patients were treated in CCU, accumulating 238 bed days. The nature of treatment is shown in Table 2. From Table 2, it can be seen that each patient treated by the CCU is likely to have an average of approximately two organs supported. The bedside electricity required to treat the patients throughout the model period depended on the machines

Table 2 e Equipment used in the CCU. Equipment name

Area

Rating (W)

Dehumidifier Feed Pump Patient statistics Touch Screen Monitor #1 Touch Screen Monitor #2 Renal Dialysis Machine Respiratory Drug Infuser Ventilator

Remote Bedside Bedside

220 90

Computing Medical areas Administrative areas Lighting Bed Bay Corridor Admin TV Nurses Station

Remote

Bedside Bedside Bedside

Remote

Discussion The model closely matches the measured energy consumption in the CCU. A small error was recorded, which predominantly reflects two reasons: 1. The model employs bed days as the unit for resource modelling. Care pathways for individual patients are assumed to last for complete 24 hours cycles, which may not be true for patients. 2. Machine use is generically applied to care pathways. Inevitably, there will be some variation in the proportion of time that machines are employed within each care pathway. It is believed that this study demonstrates that it is possible to predict electricity consumption within a CCU using this ‘bottom-up’ approach based on patient casemix. This approach can support a future sensitivity analysis around different ways of delivering care whilst maintaining clinical safety.

Table 3 e The support in days provided in the CCU and the average number of organs supported per bed day. Organ supported

20 44 200 35 125

736 1344

Remote

selected for treatment. For illustrative purposes, Table 3 contains the calculation of Eb for patients undergoing Level 3 care. The carbon emissions are estimated using a conversion factor of 0.44548 kg CO2 (eq) per kWh electricity.13 Given that the total electricity use within the Unit during the model period stood at 3600 kWh, according to the model, almost one third of the total electricity required by the Unit was used in delivering Level 3 care. Levels of electricity consumed within the Unit for treating patients of lower acuity and for non-clinical purposes are shown in Table 4. Overall the total metered electricity was 3600 kWh against a modelled total of 3636 kWh; a model error of just 1% (see Table 5). In addition to labelling energy consumption as ‘bedside’ or ‘remote’ it has been attempted to elicit the function that it is used for. For example, medical devices are ‘Clinical’, whilst information technology, when used for monitoring patients, is termed ‘Bedside IT’. Computer equipment employed for supporting duties, for example by medical secretaries, is termed ‘Administrative IT’. Fig. 1 illustrates the purpose for which electricity was used in CCU during the data period.

140 364 294 19

Level 3 Level 2 Level 1 Level 0 Total

Lung (Invasive 99 Ventilation) Lung (Non-invasive 10 Ventilation) Kidney (Renal 14 Replacement) Gut (Nutrition) 83 Heart (Vaso-active 38 drugs) Brain (Sedation) 64 Total Organ Support 308 Bed days occupied Organs supported per bed day

53

0

0

151

5

0

0

16

7

0

0

21

44 20

0 0

0 0

127 58

33 162

0 0

0 0

97 470 238 1.97

774

p u b l i c h e a l t h 1 2 8 ( 2 0 1 4 ) 7 7 1 e7 7 6

Table 4 e A derivation of the Bedside Electricity (Eb), except televisions, required in treating Level 3 patients within CCU at Treliske Hospital.

Critical Care Carbon Footprint (CCCF) Model Period Start

14/05/2012

Period End

11/06/2012

Organ Supported

Lung (Invasive)

Lung (Noninvasive)

Kidney

Gut

Heart

Brain

TOTAL

Total stay (hours)

2376

240

336

1992

912

1536

7392

Ventilator

Ventilator

Dialysis Machine

Feed pump

Spacer

Spacer

0.125

0.125

0.2

0.09

0.035

0.035

0

0

0

0

0

0

Percentage of time when on standby

0%

0%

0%

0%

0%

0%

Machine #1 kWh

297

30

67

179

32

54

None

None

None

Spacer

None

None

Loading (kW) on

0

0

0

0.035

0

0

Energy consumption (kW) when on standby

0

0

0

0

0

0

Percentage of time when on standby

0%

0%

0%

0%

0%

0%

Machine #2 kWh

0

0

0

70

0

0

70

297

30

67

249

32

54

729

142

14

32

119

15

26

349

Machine # 1 Loading (kW) on Energy consumption (kW) when on standby

Machine # 2

Total

kWh kg CO2 (eq)

659

Bedside IT per patient Monitor

Touch screen treatment applicator

Loading (kW) on

0.04

0.022

Hours Used

3743

3743

165

75

176

416

79

36

84

199

kWh kg CO2 (eq)

1145

Machine

Bedside IT kWh kg CO2 (eq)

Total Bedside Electricity Consumption for Level 3 Patients

Although the total energy modelled closely matches the total electricity expended within CCU, the modelled split between ‘bedside’ and ‘remote’ energy is a less good fit. The model of bedside energy is 224 kWh (12%) lower than metered bedside supplies while the remote energy is modelled as being 270 kWh (16%) higher than metered supplies. This may reflect ‘remote’ equipment being connected to the separate bedside electricity supply. Because it is extremely difficult to safeguard against all staff plugging equipment into the wrong circuit,

Computing (shared resource)

548

future modelling would prove more accurate in using total electricity as a quantum, rather than by modelling the bedside and remote separately. The model is useful for planning reductions in electricity usage and expenditure, valuable in helping healthcare providers reduce costs and meet carbon reduction targets.4 For a CCU to contribute to these reductions, planners need to understand the mix of patients treated in such a unit. Future energy consumption can then be extrapolated from current

p u b l i c h e a l t h 1 2 8 ( 2 0 1 4 ) 7 7 1 e7 7 6

Table 5 e Comparisons of modelled energy demand and actual energy use. Care level Bedside Energy Level 3 Level 2 Level 1 Level 0 Televisions Empty Bed Total Eb Remote Energy Lighting Other Total Er Total Used

Modelled energy (kWh)

Metered energy (kWh)

1145 410 0 0 128 10 1694

1928

838 1104 1942 3636

1672 3600

usage and literature detailing anticipated trends and the benefits of taking remedial action, including potential cobenefits, (e.g. reduced air pollution).1,6,14e17 The success of this model relies largely on the classification of patients according to organs supported through ‘levels of care’. Because a single patient may receive support for two or more organs simultaneously, it is more complex to use this methodology to identify exact resource savings within a CCU by moving a certain level of organ support away from the unit. Identification of patients needing single organ support would allow the model to identify the electricity saved to the Unit alone through an alternative treatment location for these patients if appropriate facilities were available elsewhere. However, if the patient remained in the same Healthcare Provider, the electricity consumed by the Provider in delivering care to that patient would not change unless there would be costs incurred in transporting the patient and reestablishing the patient in a new location. In moving towards the quantification of the carbon footprint of a CCU it is necessary to first gauge the proportion of energy use in a building that is taken up by electricity consumption. Applying the appropriate conversion rate for 2012 of 0.4602 kg CO2 eq/kWh for electricity generation and 0.03634 kg CO2 eq/kWh for losses in transmission and

Fig. 1 e Electricity usage in CCU during time of study.

775

distribution13 to the electricity metered through the model period, the carbon footprint of the electricity consumed per bed day in CCU is estimated to be 7.01 kg CO2 equivalent for electricity generation and 0.55 kg CO2 equivalent for transmission and distribution losses. However, in addition to the electricity consumed on CCU, additional GHG are generated in supplying hot water to wards of between one and two kg CO2 equivalent per day, based on assumptions around quantity supplied, storage temperature and water usage,18 bring the total emissions per bed day in CCU to approximately 9 kg CO2 equivalent. Consideration of heating might potentially bring this closer to that reported by Connor8 who assigned 12.4 kg CO2 equivalent from buildings use, with his calculation including heating and hot water. His study was conducted in a renal dialysis unit, where the dialysis machines and associated water and heating systems are high consumers of electricity. Other authors estimate the carbon footprint per hospital bed day to be substantially higher, e.g. 80 kg CO2 equivalent,19 but this includes factors such as travel and procurement and its indirect emissions. The model cannot as yet include the wider indirect contributions to carbon emissions, such as, for example, the 22% of emissions generated by the NHS attributable to drugs, of which critical care is a significant consumer.20 Additionally absent from the model are the emissions generated by transporting patients to and from the Unit (by land transport or air) and the transportation of commodities vital for the care of individual patients such as organs for transplant, rare bloods or rare medical devices. The model assumes that an itemized programme of care is applied uniformly to patients with the same levels of organ support. In addition to the variation in need for medical devices between identically classified patients as time passes, the need for treatment by the same patient under a constant classification is assumed to remain constant. This uniformity of use assumption represents a potential source of error as does the assumed accuracy of the electrical rating of each piece of equipment used, which was not tested. This model is bottom-up in construction, similar to that used in Pollard et al.7 Therefore, the model would be able to demonstrate how changing a care pathway might alter the emissions generated. The model is broadly consistent with the protocols recommended by the World Business Council for Sustainable Development/World Resources Institute.21 The scope of this approach is limited to the indirect GHG emissions of electricity generated elsewhere. Embodied emissions are not considered; waste is also not considered. A model has been built that is able to estimate the electricity required in a resource intensive area of healthcare, critical care, specifically the patient-facing components. This is done using patient volumes and consideration of the detail within care pathways and may be extended to other arenas where electrical devices are heavily used in care delivery. The model may also be used with patient data gathered across a longer period, for example a year, thereby quantifying the variability in consumption due to time of year or patient casemix on the Unit. The method of patient classification employed by the modelling process is based around the organ supported although the significance of using other modes of patient classification might be sought in future work. Caution is clearly needed in the use of carbon footprinting to meet

776

p u b l i c h e a l t h 1 2 8 ( 2 0 1 4 ) 7 7 1 e7 7 6

targets for carbon reduction. The potential gains from exporting emissions from CCU to other departments may make a CCU appear less carbon intensive, but lead to potentially inefficient service provision. Climate change poses both threats and opportunities to healthcare. Appropriate use of bottom-up models, such as that presented in this paper can provide significant input to decision making.

Author statements The authors are grateful for comments from Dr Paul Upton, Medical Director, Royal Cornwall Hospitals NHS Trust. All errors are our own.

Ethical approval None sought.

Funding The European Centre for Environment and Human Health (part of the University of Exeter Medical School) is supported by investment from the European Regional Development Fund 2007 to 2013 and European Social Fund Convergence Programme for Cornwall and the Isles of Scilly.

Competing interests There are no competing interests.

references

1. Costello A, Abbas M, Allen A, Ball S, Bell S, Bellamy R, Friel S, Groce N, Johnson A, Kett M, Lee M, Levy C, Maslin M, McCoy D, McGuire B, Montgomery H, Napier D, Pagel C, Patel J, de Oliveira JA, Redcliff N, Rees H, Rogger D, Scott J, Stephenson J, Twigg J, Wolff J, Patterson C. Managing the health effects of climate change: Lancet and University College London Institute for Global Health Commission. Lancet 2009;373(9676):1693e733. 2. Godlee F. How on earth do we combat climate change? BMJ 2011;343. 3. Chung JW, Meltzer DO. Estimate of the carbon footprint of the US health care sector. JAMA 2009;302(18):1970e2. 4. Cole A. More treatment in surgeries and at home will help cut NHS carbon footprint. BMJ 2009;338:b345. 5. Sustainable Development Unit. Saving carbon, improving health: update carbon reduction strategy 2010. Available at: http://www. sdu.nhs.uk/documents/publications/UPDATE_NHS_Carbon_ Reduction_Strategy_(web).pdf; 2010 (accessed 26 July 2011).

6. Charlesworth A, Gray A, Pencheon D, Stern N. Assessing the health benefits of tackling climate change. BMJ 2011;343:d6520. 7. Pollard AS, Taylor TJ, Fleming LE, Stahl-Timmins W, Depledge MH, Osborne NJ. Mainstreaming carbon management in healthcare systems: a bottom-up modeling approach. Environ Sci Technol 2013;47(2):678e86. 8. Connor A, Lillywhite R, Cooke MW. The carbon footprint of a renal service in the United Kingdom. QJM Mon J Assoc Physicians 2010;103(12):965e75. 9. Chapman M, Chapman A. Greening critical care. Crit Care 2011;15(2):302. 10. Royal Cornwal Hospitals NHS Trust. About us. Available at: http://www.rcht.nhs.uk/RoyalCornwallHospitalsTrust/Our Organisation/AboutUs.aspx; 2008 (accessed 19 July 2011). 11. Cornwall Council. Population data for Cornwall. Available at: http://www.cornwall.gov.uk/default.aspx?page¼22123; 2011 (accessed 20 July 2011). 12. Department of Health. Comprehensive critical care: a review of adult critical care services 25/5/2000. Available at: http:// webarchive.nationalarchives.gov.uk/þ/www.dh.gov.uk/en/ PublicationsAndStatistics/Publications/PublicationsPolicy AndGuidance/DH_4006585; 2000 (accessed 5 February 2013). 13. Defra. 2014 Greenhouse gas conversion factor repository. Available at: http://www.ukconversionfactorscarbonsmart. co.uk/LandingPage.aspx; 2014 (accessed 08 May 2014). 14. McMichael AJ, Neira M, Bertollini R, Campbell-Lendrum D, Hales S. Climate change: a time of need and opportunity for the health sector. Lancet 2009;374(9707):2123e5. 15. Haines A, McMichael AJ, Smith KR, Roberts I, Woodcock J, Markandya A, Armstrong BG, Campbell-Lendrum D, Dangour AD, Davies M, Bruce N, Tonne C, Barrett M, Wilkinson P. Public health benefits of strategies to reduce greenhouse-gas emissions: overview and implications for policy makers. Lancet 2009;374(9707):2104e14. 16. Woodcock J, Edwards P, Tonne C, Armstrong BG, Ashiru O, Banister D, Beevers S, Chalabi Z, Chowdury Z, Cohen A, Franco O, Haines A, Hickman R, Lindsay G, Mittal I, Mohan D, Tiwari G, Woodward A, Roberts I. Public health benefits of strategies to reduce greenhouse-gas emissions: urban land transport. Lancet 2009;374(9705):1930e43. 17. Patz JA, Campbell-Lendrum D, Holloway T, Foley JA. Impact of regional climate change on human health. Nature 2005;438(7066):310e7. 18. Absolicon. Hospital solar energy 2011. Available at: http://www. absolicon.com/_photovoltaic_thermal_pvt/1042_hospitals_ eng.php; 2011 (accessed 27 November 2011). 19. Reynolds L. Climate change is recognised as a threat to human health. Continuing Medical Education, http://cmej.org.za/ index.php/cmej/article/view/2347/2208; 2012 (accessed 01 May 2012). 20. Moynihan R. The greening of medicine. BMJ 2012;344:d8360. 21. Bhatia P, Ranganathan J. The greenhouse gas protocol: a corporate accounting and reporting standard (Revised Edition) March 2004, http://www.wri.org/publication/greenhouse-gas-protocolcorporate-accounting-and-reporting-standard-revisededition; 2004 (accessed 31 August 2012).

The carbon footprint of acute care: how energy intensive is critical care?

Climate change has the potential to threaten human health and the environment. Managers in healthcare systems face significant challenges to balance c...
411KB Sizes 1 Downloads 3 Views