Development of an Automatic Heel Ulcer Prevention Mattress

(1)

Development of an Automatic Heel Ulcer Prevention Mattress

LINUS GREEN

Master of Science Thesis Stockholm, Sweden 2016

(2)

Development of an Automatic Heel Ulcer Prevention Mattress

Linus Green

Master of Science Thesis MMK 2016:52 IDE 151 KTH Industrial Engineering and Management

Machine Design SE-100 44 STOCKHOLM

(3)

ii

(4)

iii

Examensarbete MMK 2016:52 IDE 151

Utveckling av en automatisk madrass för förhindrande av trycksår på hälar

Linus Green

Godkänt

2008-mån-dag

Examinator

Claes Tisell

Handledare

Teo Enlund

Uppdragsgivare

Jack Spira

Kontaktperson

Jack Spira

Sammanfattning

Detta examensarbete är en del i ett produktutvecklingsprojekt som initierats av Jack Spira. Dess syfte har varit att leverera ett koncept som automatiskt förhindrar trycksår på hälen hosvårdpatienter. Detta koncept kommer sedan utvecklas vidare i ett startup-företag och sedan förhoppningsvis integreras i vården.

För att kunna utveckla ett koncept som förhindrar, eller drastiskt reducerar antalet, trycksår på hälar gjordes en grundlig förstudie. Studien inkluderar litteraturstudier för att förstå orsaker och mekanismer bakom trycksår samt en genomgång av befintliga behandlingar och förebyggande åtgärder. Efter en idégenereringsprocess bedömdes, diskuterades och sållades de olika idéerna inom projektgruppen.

Det slutgiltiga konceptet är en madrass, kompatibel med befintliga vårdsängar, som med hjälp av sensorteknik och pneumatic helt avlastar patientens hälar och därmed förhindrar uppkomst av trycksår. Förfarandet är helautomatiserat för att utesluta den mänskliga faktorn och det behov av tidig upptäckt som finns i dagens metoder.

Konceptet kommer att utvecklas vidare av det startup-företag som bildats av projektgruppen. Framtida planer innefattar bland annat prototyptillverkning för slutgiltig teknisk utvärdering och en klinisk studie för att kvantifiera dess potentiella inverkan på antalet trycksår som uppstår.

(5)

iv

(6)

v

Master of Science Thesis MMK 2016:52 IDE 151

Development of an Automatic Heel Ulcer Prevention Mattress

Linus Green

Approved

2008-month-day

Examiner

Claes Tisell

Supervisor

Teo Enlund

Commissioner

Jack Spira

Contact person Jack Spira

Abstract

This thesis is part of a product development project initiated by Jack Spira.

The purpose was to deliver a concept that would automatically prevent pressure ulcers on the heels of hospitalized patients. This concept would then be realized through a startup project and hopefully integrated in healthcare worldwide.

In order to develop a concept that eliminates or drastically restricts the prevalence of heel ulcers a thorough background study was conducted. The study included reviewing the causes and effects of pressure ulcers as well as current treatments and preventive measures. In the ideation phase ideas were generated and then evaluated and screened with input from other project members.

The resulting concept is a mattress, integratable in existing ward beds, that uses a censoring technique and pneumatics to completely unload the heels and thus eliminates the risk of heel ulcers. The process is completely automated to prevent human error and the current need for early detection.

The concept will be further developed by the startup company sprouting from the project. The process will include prototyping and testing to fine tune the function as well as clinical trials to quantify the potential effect on heel ulcer prevalence in the healthcare system.

(7)

vi

Abbreviation dictionary

Pressure Ulcer/s ... PU/s European pressure ulcer advisory panel ... epuap National pressure ulcer advisory panel ... npuap point prevalence study ... PPS Royal Institute of Technology (Kunliga Tekniska Högskolan) ... KTH Pairwise Comparison Chart ... PCC Alternating Pressure Air Mattress ... APAM Unique Selling Point ... USP

(8)

(9)

1

1 Introduction

This thesis was part of an ongoing project that was initiated shortly before the involvement of KTH. The project was initiated by Jack Spira, who is a MD. PhD. Docent in tumor biology and currently the CEO of a private medtech company, after a friend of his suffered from heel ulcers. Later on Martin Löfbom, MSc. in industrial economy and business architect at

Flemingsberg Science, was recruited along with Eila Sterner, nurse and MD.

PhD in pressure ulcer prevention. Via the Flemingsberg Science network KTH was involved, eventually resulting in the initiation of this thesis. All of the above are referred to simply as the project group throughout this paper.

(12)

4

(13)

5

2 Background

This chapter defines pressure ulcers, as well as their known causes and the problems they currently cause in health care. The main focus will be

pressure ulcers located on the heels as those are the focus of the project.

2.1 Definition

Pressure ulcers, also known as bedsores, pressure sores and decubitus ulcers, are ulcerations of skin and underlying tissue caused by pressure.

Persons who have been bedridden for a long time are at bigger risk of developing pressure ulcers and they are most likely to appear over bony prominences (Britannica, 2015). Early discovery is crucial as PUs are

increasingly hard to heal the further they are allowed to progress. In severe cases, usually regarding PUs below the knee, amputation may be required (Evonne Fowler, Suzy Scott-Williams, & and James B. McGuire, 2008).

The European pressure ulcer advisory panel (epuap), together with the National pressure ulcer advisory panel (npuap, USA), have created a quick reference guide for the prevention and treatment of PUs. In this document PUs are split into four main categories based on their severity, ranging from low to high. This definition is called the international npuap/epuap pressure ulcer classification system. Figure 1 is an excerpt from the document,

showing the definitions of category I to category IV.

(14)

6

Figure 1, the international npuap/epuap pressure ulcer classification system (I-IV) (National Pressure Ulcer Advisory Panel, 2014)

There are two additional categories for when the severity is undefinable, however, these are uncommon in Europe and therefore of little relevance to this thesis.

2.2 Causes

Ischemia (inadequate blood supply) is one of the factors causing damage to tissues under sustained pressure. Tissue under pressure will also be

subjected to hypoxia, blocking of nutrient supply and blocking of the removal of waste products (National Pressure Ulcer Advisory Panel E. P., 2014). In a study by Sangeorzan et al. the ischemic effects are tied to pressure by measuring the transcutaneous partial pressure of oxygen

(15)

7 (TcPO2) while applying a variety of external loads. Measurements were made on skin covering muscle tissue and covering bone. TcPO2 levels indicate the amount of oxygen being supplied to the skin tissue. In this study TcPO2=0 was used as a threshold, that is when all of the supplied oxygen is consumed in the metabolism of the cells and thus the threshold for ishemia (Bruce J. Sangeorzan, 1989). The results showed that skin covering bone is the most sensitive to external pressure, reaching TcPO2=0 at pressures of around 40 mmHg (5,3 kPa) as shown in figure 2.

Figure 2, Pressure and displacement measurements for when TcPO2=0 over the tibia bone and tibialis anterior muscle. (Bruce J. Sangeorzan, 1989)

Shear forces and frictional forces are also commonly present meaning that the total load affecting tissues exceeds that of pressure alone (Ryan P . Crenshaw, 1989). The sum of these forces is sometimes referred to as the mechanical load (National Pressure Ulcer Advisory Panel E. P., 2014). When tissues are under sustained mechanical load pressure ulcers have been shown to appear within time frames ranging from 2 to 6 hours (Kosiak, 1959).

Another factor is damage directly induced by deformation, as shown in a recent study by Gefen et al. In this study a graph was created showing how different strains cause cell death in muscle cells over various time intervals, which is shown in figure 3. The study also shows that high strains (above 50% true compressive strain) can cause damage to muscle cells in a matter of minutes (Amit Gefen, 2008).

(16)

8

Figure 3, graph showing the relation between strain and time in muscle cells (Amit Gefen, 2008)

Other factors related to the likelihood of developing a PU include age, nutrition and dehydration (Improvement, 2011). These and other patient specific factors affect the tolerance of the skin to mechanical load and thus the risk of developing them.

A tool widely used for the assessment of patient risk is the Braden scale. The scale includes six subcategories related to PU risk which are “sensory

perception”, “moisture”, “activity”, “mobility”, “nutrition” and “shear and friction” (Evonne Fowler, Suzy Scott-Williams, & and James B. McGuire, 2008). An example of a Braden scale assessment chart is shown in figure 4.

(17)

9 Figure 4, A Braden scale assessment chart (Barbara M. Bates-Jensen, 2001)

Based on the received score caregivers can evaluate the need for preventive measures to be used for each individual patient. A lower score means higher risk and bigger need for specialized preventive equipment and/or special meals, skin care etcetera.

2.3 Heel ulcers

PUs located on the heel are sometimes referred to as heel ulcers. Roughly one third of all PUs are heel ulcers. Pain and impaired mobility are common problems, but heel ulcers can also lead to loss of limb, through amputation.

A prospective study of patients with heel ulcers found that 11 out of 154 participants were amputated despite receiving care (Evonne Fowler, Suzy Scott-Williams, & and James B. McGuire, 2008). Apart from the loss of a limb a study of 788 patients with lower limb amputations showed a re- operation rate of 18.4%, a mortality rate of 5.7% and an infection rate of 5.5% (Bernadette Aulivola, o.a., 2004) emphasizing the risks for patients with heel ulcers.

(18)

10

The prevalence varies across patient populations but gets as high as 25.1% in care settings. Suffering from a PU has been shown to almost

double the morbidity in elderly patients. PUs affecting the lower extremities, roughly one third of the total number, are the most complicated ones to treat. In the worst cases these can lead to the need for a wheelchair or even amputation.

The costs of treating a single pressure ulcer is hard to define but can range from around a thousand USD to more than ten thousand USD (Evonne Fowler, Suzy Scott-Williams, & and James B. McGuire, 2008).

2.4 Problem

PUs are an ongoing issue in Swedish healthcare today. Despite the preventive methods and products in use the prevalence remains fairly stable, as shown in a point prevalence study (PPS) by “Sveriges kommuner och landsting”. In this study the average prevalence of PUs in inpatient care in Sweden was found to be 13.6% (Landsting, 2015). The prevalence varies over patient groups, with ages 80 and above showing the highest numbers at around 20%. The same study shows that, of those admitted, 19.4% of patients are considered at high risk of developing PUs. Of those considered at high risk, 36.3% will develop PUs despite the use of preventive measures in 84.2% of their cases. The study also shows what preventive measures are taken into account, as well as their individual prevalence, as shown in figure 5.

(19)

11 Figure 5, Preventive measures used. (Landsting, 2015)

The preventive measures listed, here sorted by frequency high to low, are:

-Specialized mattresses -Scheduled repositioning -Heel elevation

-Sliding sheets

-Scheduled repositioning in chair -Pressure minimizing cushion in chair -Other

(20)

12

(21)

13

3 Project task

The task for the project, of which this thesis is a part, is to develop a new preventive product for heel ulcers. Since other such products have been shown to reduce the incidence of heel ulcers (Donnelly, Winder, Kernohan, &

Stevenson, 2011) it has been assumed that human error e.g. regarding early discovery, and financial restrictions in healthcare are partial causes to the unchanged prevalence. One key aspect of the project is therefore that the product should be integrated in the hospital bed and function

automatically, thus eliminating human error while hopefully significantly decreasing the prevalence of heel ulcers. A requirement specification was made early in the process to clarify the expectations of the final concept.

The requirement specification is shown in figure 6.

Figure 6, requirement specification

Apart from integration and automatic pressure relieving sterilization and independence of patient size are important requirements. Sterilization is crucial, as products in healthcare environments must meet the required standards to be allowed. Independence of patient size is of high importance since the product should be integrated into the bed and automatic. Therefore it cannot, and should not have to, be tapered to individual patients. Being able to work with each leg separately is useful e.g. if one leg has been fractured and is in a cast.

It was decided that ward beds should be in focus and that operation beds should be excluded since they differ too much in their design to share the same integratable product. Additionally, fewer patients lie in an operating bed, and for less time.

(22)

14

The main deliverable of the thesis is a finalized concept meeting the

specified requirements. It was decided in the early stages that no prototype should be included since it would likely be made with methods, tools and materials inaccessible to students at the Royal Institute of Technology (KTH).

3.1 Initial concept

Together with the thesis proposal came an initial concept, which had been part of the spark that started the project and was meant to be the basis of the thesis work. The idea was based on existing PU prevention air

mattresses which inflate and deflate part of the surface to relieve pressure on set intervals of time. The idea was to use the same principle but scale it down so as to better protect the heels. This downscaled system would then be combined with a sensor system that would monitor the patient's

movement and activate the preventive measures when needed.

With this idea in mind the original thesis proposal was for two students, one from integrated product design and one from the mechatronics department, to make a finalized concept. However, since only one student was recruited it was decided that the task should be expanded in the product development area to include background research and idea generation that would

potentially develop the idea further before conceptualizing it.

(23)

15

4 Method

The following describes the plan on which the thesis part of this project will be based.

4.1 Process

Steps included in the method were generally defined in the project proposal as pre-study, design proposal and prototyping. However, as mentioned before, as the task was redesigned to fit a single student the proposed process was also redefined. In the final version prototyping would be replaced by a more substantial research phase and idea generation. To fit the new task the pre-study would have to include a more detailed research on PUs, to identify possible ways of preventing them. It would also have to include a competitor analysis to identify and possibly draw inspiration from products currently available. The idea generation phase sought to

complement to the initial concept while providing a broader perspective.

Upon completion of the idea generation phase the entire group therefore evaluated all ideas and decided which one to conceptualize. The finalized concept should be specified enough for a prototyping company to build a concept based on the details provided. Mock-ups may be created to illustrate the size, shape and simplified function of the concept.

A process plan was created for the project and is shown in figure 7. It is based on the model defined in “The Mechanical Design Process” (Ullman, 2010) but has been modified to fit the project at hand. Changes were made in order to compensate for parts being pre-defined, such as customer

identification and need.

Figure 7, process plan

4.2 Time plan

Based on the information at hand at the start of the project a rough time plan was created. Keeping in mind that process steps and deliverables may change over the course of the project it was used as a schedule and an overview, especially during the initial phases. A phase of particular uncertainty was the background study, as it was not yet known whether

(24)

16

relevant information would be found or additional experimenting needed. A Gantt schedule was created and is shown in figure 8.

Not featured in the figure are the regular checkup-meetings that will be scheduled once every other week throughout the project, enabling all involved to follow the progress being made.

Figure 8, Time plan

4.3 Methodologies

This chapter explains methodologies used that will be referred to in later parts.

4.3.1 Pairwise comparison chart

When weighting the requirements specified for a project a pairwise

comparison chart (PCC) can be used. When utilizing the method of pairwise comparison, each alternative is compared one-on-one with each of the other alternatives. Alternatives are rewarded one point for each win, half a point for each tie and zero points for each defeat. Then the total score is

calculated and determines the ranking of all alternatives (The Method of Pairwise Comparisons). An example of a PCC is shown in figure 9. When one or more requirements get a score of zero, adding one to all scores enables use in e.g. a Pugh’s decision matrix.

Figure 9, example of a PCC.

(25)

17 4.3.2 Pugh’s matrix

A Pugh’s matrix is used for screening alternatives and decision making. The basic idea is to evaluate how well each concept fulfills each requirement. By also including the weight factor of each requirement an overall score can be given, enabling comparison of concepts that are good based on different aspects (Ullman, 2010). The structure of a Pugh’s matrix can be seen in figure 10.

Figure 10, structure of a decision matrix.

One way of scoring the concepts is using one of them as reference.

Remaining concepts then receive +1, -1 or 0 depending on whether they are considered to be better, worse or equal to the reference. Each score is then multiplied by the weight factor of its corresponding requirement and

summed up to a final concept score. Using this scoring method the reference will always get 0 points, while remaining concepts can receive with both positive and negative scores.

(26)

18

(27)

19

5 Implementation

This chapter will provide description of steps that were taken and proved relevant to the resulting concept. It includes background studies, ideation and development process of the final concept.

5.1 Market screening

A market screening was conducted in which existing products related to pressure ulcer prevention and pressure monitoring were reviewed. As many of these overlap, a summarized selection will be shown here to briefly

present the different approaches currently available.

5.1.1 Static solutions

Static solutions are products which are passive. Where special products are not available these sometimes include pillows or water-filled rubber gloves to decrease the pressure on a patient's heel. The general principle of

prevention is either cushioning or completely unloading a sensitive area by lifting it off the surface.

There are static products specifically designed to prevent pressure ulcers on the patient's feet. A common design among these is a boot made either of polymeric material with air-filled cells or foam, elevating and unloading the heel when strapped to the foot. An example of such a product is shown in figure 11.

Figure 11, Heellift boot. (DM Systems, Inc., 2008)

(28)

20

Another approach is adhesive padding, similar to a band aid, that is applied to a specific location either as a preventive measure or to protect an ulcer that has already began to form. These pads come in various shapes and some have gel internals to relieve pressure or medicine to help heal existing wounds. An example of such a product is shown in figure 12.

Figure 12, adhesive heel padding. (Wound Care Today, 2017)

Some companies target all areas sensitive to pressure ulcers by providing special mattresses designed to maintain low pressures and manage

microclimate. This is commonly achieved using memory foam combined with multiple sub layers that transfer moisture away from the patient while

distributing pressure across the largest possible area. Some of these also feature a sloped section at the lower end meant to decrease the pressure applied to the feet. One such mattress is shown in figure 13.

Figure 13, low pressure mattress. (Care of Sweden AB, 2016)

(29)

21 5.1.2 active solutions

Products that consume energy to actively redistribute pressure on the patient's body, usually based on predetermined intervals of time, are here categorized as active solutions. The most basic of active prevention is

scheduled repositioning, where caring staff reposition the patient in order to relieve pressure from altering parts of the body. Various products that aid this process are available, an example is sliding sheets that are placed under the patients to minimize friction and enable caring staff to rotate the patient without lifting them off the surface of the bed.

There are several products available that essentially consist of several inflatable tubes stacked side by side to the full length of the bed. These tubes control which parts of the body are submitted to pressure by deflating different sets of cells over time, usually in a pattern like every other or third tube. This allows for tissue to recover while the tube with which it is in

contact is deflated. One such mattress is shown in figure 14.

Figure 14, Alternating pressure air mattress. (Care of Sweden AB, 2016)

5.1.3 pressure monitoring

One current monitoring system maps the pressure distribution on bedridden patients by combining measurements from several sensors on a sheet-like overlay and displaying them as a pressure map on a display. This has been used when positioning patients to make sure that correct postures are used where no point on the body is exposed to pressure exceeding a set

threshold. It may also be used to evaluate the performance of pressure ulcer prevention mattresses (Tekscan, Inc., 2016).

(30)

22

5.2 Literature review

A literature review was conducted in order to gain understanding as to the causes and mechanisms behind the development of PUs. Most of the findings are covered in the background chapter of this paper, as it concerns the

definition, causes and prevalence of PUs. Other aspects studied are covered here.

With the initial concept in mind, two key variables were identified and sought to be defined. These were threshold values for pressure and time, since these were the factors to be manipulated. Another aspect was defining some sort of minimum working area, to determine whether altering pressure could be done at a micro scale, barely noticeable to the patient.

It turned out that though much is known about PUs and what causes them, identifying the importance of a single contributing factor is hard and varies between individuals. For example, it is known that a multitude of factors, such as diet, age and mobility, contribute to the skin’s tolerance to pressure (Improvement, 2011). However, none of the reviewed papers has been able to isolate a single factor and determine its influence on PUs. Because of this, the thresholds of time and pressure vary significantly between patients and exact values could not be found.

The review also covered studies evaluating the efficiency of various existing solutions. These include heel boots, alternating pressure air mattresses (APAMs) and seat cushions for wheelchairs. Some of the results found in these studies proved to be of value when evaluating ideas and are thus covered here.

In a study by Shyam et al. four different British made APAMs were evaluated. As part of the evaluation the interface pressure (IP), between different parts of the patient's body and the mattress, was monitored (Shyam V.S. Rithalia, 1998). The table showing the resulting pressures and standard deviations is shown in figure 15.

Figure 15, IP results on specific locations (Shyam V.S. Rithalia, 1998)

As shown in figure 15 the mean maximum values on the heel range from 147.1mmHg to 186.9 mmHg. These values can be compared to the values

(31)

23 found by Sangeorzan et al. in figure 2 (background chapter). These show the IP for which TcPO2=0 over the tibia and tibialis anterior muscle, ~40 mmHg and ~70 mmHg respectively. As there is no muscle tissue at the heel the IP over the tibia (40 mmHg) is the most relevant comparison. It should be noted that TcPO2=0 is not a threshold for pressure causing PUs.

However, it does specify the threshold for when all available oxygen is consumed by the cells, and thus indicates the point beyond which ishemia could be an issue.

Additionally, another study by Vanderwee et al. found that the incidence of PUs was comparable in patients nursed on an APAM and those nursed on visco-elastic mattresses and turned every four hours (KATRIEN

VANDERWEE, 2005). Consequently, it was assumed that APAMs do not prevent PUs on the heel.

In a study by Donnelly et al. the effectiveness of heel offloading devices, such as boots, compared to standard care was examined (Donnelly, Winder, Kernohan, & Stevenson, 2011). All 239 subjects were over 65 years of age and had suffered a hip fracture, 119 of these were assigned to the control group and 120 to the intervention group. Patients in both groups were nursed on pressure-redestributing surfaces, both static and active, as assigned by the staff based on individual need. The patients in the

intervention group were also equipped with boots unloading the heel, on both feet. The study found that none of the subjects in the intervention group developed any PUs on their ankles, feet or heels while 29 subjects in the control group developed PUs in these areas. The total occurrence of PUs was 26% in the control group and 7% in the intervention group. Another relevant finding was that though the boot seemed to prevent PUs not all patients were pleased with its comfortability.

5.3 Visit at the Karolinska university hospital

To better understand the environment in which the product is meant to be used a visit at Karolinska was arranged. Input was gathered by auscultating an assistant nurse for two hours. The preset was that nursing routines and conditions were part of the problem, as it has been shown that existing products can prevent PUs. The focus was thus on the caring environment as a whole, regarding eg. stress levels and PU routines, as well as available equipment and its use.

Unfortunately the visit was scheduled on a Friday afternoon, resulting in there being unusually few patients admitted. Additionally, none of the seven patients admitted at the time had any developed PUs. Another circumstance that limited the gathering of information was that the assistant nurse being auscultated had only been working for four months and therefore lacked experience regarding some working routines and equipment.

A few conclusions could be drawn from the interview regardless. The standard practice, for preventing heel ulcers, at the visited ward, was the use of thin foam pads that were placed under the calf of the patient to elevate the heel. The general opinion of the local staff however was that

(32)

24

these pads were not very effective. The first main contributing factor to their conclusion was that the pad was too thin and they had to stack two of them on top of each other to unload the heel. The second was that the pads did not stay where they were supposed to if the patient moved at all. Meanwhile a large quantity of the “lassekudde”, a type of foam boot which presumably solves both issues mentioned (Solann AB, 2016), were available. Brief

contact was made with a nurse responsible for pressure ulcer prevention and when asked why “lassekudden” was not used she replied that it should be, and that early signs of ulceration had been detected on a number of

admitted patients. Evidently though, some of the staff did not know what

“lassekudden” was, or how it should be used. This was of course an isolated case and does not necessarily apply in general.

It was also noted that APAMs were in frequent use, contrary to what was expected. However, it was explained that the availability of APAMs differ a lot between wards and therefore their frequency of use was probably not that high in general.

5.4 Visit at the bed workshop

To evaluate the possibilities and restrictions of integration ward beds had to be further studied. For this Lars Halvardsson, who is responsible for bed maintenance in the workshop at Karolinska Huddinge, was contacted. A meeting was set during which he would demonstrate some of the models currently in use, as well as future ones, and answer the questions posed.

The most relevant information sought after was measurements between hinge points, where the mattress can be positioned, as well as the design and compressibility of the mattresses used. These were important aspects as they may restrict the size of the product and influence the extent to which it can be integrated into the mattress. The measurements obtained cannot serve as an exact reference however, since it became clear that beds are procured yearly to replace worn out ones. Additionally, purchases are made with regard to best price at current time, which means that there are several models in use at any given moment. Additionally, procurement is done at a local scale, in this case the hospital in Huddinge did its own. This means that the models of beds in use differ between hospitals as well as within them. A general measurement limit thus had to be defined. It was concluded that modern beds were split into four sections, as illustrated in figure 16, and that the product would have to be located only in the foot section.

(33)

25 Figure 16, generalized modern bed design.

The length of this lowest section ranged from 51 to 54 centimeters on

available models. It was thus decided that the product should not exceed 50 centimeters to avoid interference with the hinge point.

It was agreed that a discarded foam mattress could be borrowed for further tests regarding product integration. Possibly, this sample could also become part of future concepts for functionality evaluation.

5.5 Ideation

Most of the ideation process took place in parallel with the literature studies and market screening. Whenever an idea emerged on how to solve some part of the problem it was documented as a sketch. Generally, the problem was treated as two separate sub problems. The first of these was preventive activity i.e. the main function that would restrict or redistribute pressure applied to the heel. The second was monitoring i.e. the function enabling the product to decide when and how to interfere.

(34)

26

5.5.1 Preventive activity

The preventive activity can be subdivided into different approaches based on documented ideas, as shown in figure 17. Most of these categories also roughly apply to products currently on the market. Categories in the lowest levels where assigned abbreviations, to be used as reference for the

concepts.

Figure 17, problem approach subdivision chart.

Figures 18-22 show a few examples of ideas, representing each of the approaches displayed in figure 17. All sketches can be seen in appendix 3.

Figure 18, Three ideas focused on elevating the heel. (EF-1) using motorized joints, (EF-2) using a pillow able to slide along the bed and (EF-3) using inflatable cells.

(35)

27 Figure 19, Three ideas that lower the surface of the bed to unload the heel. (LS-1) compresses the mattress, (LS-2) is a cut out shape in the mattress and (LS-3) is a pillow

that can reshape itself.

Figure 20, Three ideas that reposition the foot to redistribute the pressure. (RF-1) through a rotating motion and (RF-2 & RF-3) by tilting the foot from side to side.

(36)

28

Figure 21, Two concepts that alter the surface of the bed to redistribute pressure. (AS-1) uses a low friction sheet to move the patient across a textured surface and (AS-2) uses

small air filled tubes in a wavy configuration that are deflated in sequence.

Figure 22, two ideas on distributing pressure better across the surface of the heel. (ED-1) uses air filled bubbles that redistribute a fixed volume of air as they are loaded. (ED-2) is a sheet of donut shaped air cells meant to capture the heel and better distribute the pressure.

5.5.2 Monitoring

Monitoring ideas can be subdivided into two subcategories, contact (C) and non-contact (NC). Contact sensor alternatives found and considered include pressure, heat, moisture and conductivity. Any of these sensor techniques could potentially be attached directly to the preventive activity design or to a fabric covering it. Some examples of such combinations were sketched and two are shown in figure 23.

(37)

29 Figure 23, contact sensors combined with preventive ideas. (C-1) sliding pillow with integrated contact sensor. (C-2) mattress compressing device with integrated contact

sensor.

Noncontact ideas differed a bit more in their function and a few separate ones were sketched. These can be seen below in figure 24.

Figure 24, noncontact ideas for monitoring. All three ideas featured are meant to determine the distance between the feet of the patient and the end of the mattress, but using different

techniques. (NC-1) uses ultrasound, (NC-2) uses a 3D camera and (NC-3) uses lasers.

(38)

30

As there was no interference between ideas of preventive activities and monitoring i.e. they could be freely combined, they were kept separate through the first screening phase. However, while brainstorming on

combinations two additional ideas on preventive activity arose. One was a merger of EF-1 and LS-1, named M-1. The other was a merger of EF-2 and LS-1, named M-2. These are shown in figure 25.

Figure 25, merged preventive activity ideas. (M-1) combines compression of the mattress with air cells and (M-2) combines it with an elevating pillow.

Both merged ideas have an advantage to their unmerged lifting originals in that the lift does not have to be as high compared to the surface of the bed.

This allows for a more natural supine position to be maintained.

Screening

Screening of the ideas was conducted by the entire group during a meeting.

The main screening concerned ideas on preventive activities, as it was agreed that the monitoring solution could be chosen and applied at a later stage. Sketched ideas were printed and brought to the meeting as the basis for discussion.

5.6 Gut feel screening

A screening of the ideas based on experience and gut feel was conducted during a group meeting, where all participants were present. It resulted in three concepts being chosen for further evaluation. These three top

candidates can be seen in figure 26.

(39)

31 Figure 26, top candidates from the ideation screening.

Previous names are now discarded and the concepts are henceforth referred to as concept one, two and three according to the number assigned to each in figure 26. To ease the understanding of the upcoming decision process each of the three concepts are briefly described below.

5.6.1 Concept one

The first is a cushion, roughly formed as a wedge, attached to the mattress using two straps. The cushion is motorized and can relocate itself along a wire according to the height of the patient. Covering the cushion is a sheet that minimizes the friction when sliding back and forth while providing a good surface for contact monitoring.

5.6.2 Concept two

The second concept combines compression of the mattress with air filled cells. The combination enables the cells to function in both directions

vertically, both raising the leg and lowering the surface. monitoring could be done either through contact, with sensors attached directly to the surface of the cells or to a covering fabric, or a non-contact idea.

5.6.3 Concept three

The third concept also utilizes compression of the mattress to lower the contact surface. However, in this case compression points are activated selectively based on where the foot is located. Monitoring would ideally be done through contact sensors located on the top surface, which is also the anchor for the compression straps along the sides.

(40)

32

5.7 Pairwise Comparison Chart

In preparation of deciding on a final concept a Pugh’s matrix would be used to emphasize strengths and weaknesses of the three candidates. Before this could be done, however, the requirements had to be weighted. This was done using a pairwise comparison chart (PCC). The setup for this is shown in figure 27. In the matrix the requirements are simply listed as numbers, however, thus the resulting ranking with score is shown in figure 28.

Figure 27, PCC implementation.

Figure 28, PCC results.

(41)

33 5.8 Decision matrix

Concept one, two and three were put in Pugh’s decision matrix to provide some input to the following meeting, at which a final concept would be chosen. Concept one was assigned to be the scoring reference and the remaining two were scored with +/- 1 or zero depending on whether they were better, worse or equal to the reference at fulfilling a requirement. The sum of these numbers multiplied with the weight of the corresponding requirement gave the resulting score displayed in the matrix. The matrix is shown in figure 29.

Figure 29, the completed decision matrix.

The decision matrix was dominated by zeros, which is explained by the fact that distinguishing how well a concept is able to solve a requirement was not always easy in this early stage of development. Furthermore, the nature of the requirements made ranking impossible in some cases. For example, the first requirement is “allow full integration” and all three concepts do. The lack of differences among the results led to further discussion regarding with which concept to proceed. During this conversation patentability surfaced as an additional requirement, not previously discussed.

At the next meeting, based on the combined input of the decision matrix and experience, including patenting, it was eventually agreed that concept two should be developed further. Among the key components to this decision was its ability to work in both vertical directions, allowing it to support the leg in rotation to some extent, while simultaneously unloading the heel by deflating cells. Additionally, the ability to lower the surface of the bed is unprecedented, as far as the collective knowledge of the group goes, providing leverage when applying for a patent.

(42)

34

(43)

35

6 Concept development

Having specified the outlines of the final concept, final deliverables for the thesis part were further defined. Knowing that the concept would consist of two main parts, the part integrated in the mattress and the electronics operating it, it was agreed that this thesis should be focused on defining the first. Supporting this decision was the fact that the thesis was originally meant for two students, one of which would have the experience required to work on the mechatronic part. As no student with such experience was

found in time the plan was to involve a second thesis work later if necessary, which it turned out to be. However, as the controlling unit would still be tied to the functionality of the concept, a brief defining the required functions would be a deliverable from this thesis to lay the foundation for the next. Furthermore, as it was clear at this point that a future prototype would require materials and tools beyond what was accessible through KTH, it was agreed that finalizing the concept to the point where prototyping could be outsourced was the main deliverable of this thesis. Additionally, to further utilize the skill set within the current thesis another part was added to the controller unit brief as a deliverable, concerning the user interface of the future product. In conclusion there were now two deliverables to the thesis, in ranking order as follows:

 A finalized concept, ready for prototyping

 A brief describing the functions and interface of the controller unit

6.1 Pneumatic controller idea

The first step in the finalizing development of concept two was to decide which method should be used for monitoring. In an effort to keep the part of the concept integrated into the mattress as simple as possible a new idea emerged in addition to the ones described in the previous chapter. The idea was to utilize the pneumatic system, which would be required for controlling the cells inflation, also in the monitoring function. By measuring the air pressure individually in each cell the hope was to use that information to identify the cells that were carrying load i.e. supporting the leg. The last cell showing external load would give the location of the heel, with accuracy dependent on the cell size. This theory posed several questions however, since ability to detect movement was also required to compensate for the patient shifting their position on the mattress. Additionally, some sort of reset cycle would be necessary to determine the new location of the heel after such a repositioning.

In order to gain further understanding of the potential of this idea contact was made with a former teacher at KTH, Jan D Hölcke. Currently retired, Jan had previously taught a course in pneumatics and kindly offered to provide some input at a meeting.

(44)

36

Upon explaining the task and concept different ways of monitoring using the pneumatic system were discussed. One problem profoundly discussed was determining the state of each cell i.e. its grade of inflation. This problem arises when the cells are to be reflated, which would be required to relocate the heel if the patient moves. The first spontaneous idea was monitoring the volume of air being pumped into each cell. However, this was discarded as air changes its volume when pressure is involved, which it would be when inflating a cell on which the leg was resting. Another approach was to make sure that the pump was strong enough to fully inflate the cells and just inflate maximum each time. This would cause problems with locating the heel however, since all cells would then have the same internal pressure, making it impossible to determine which were under external load. Further discussion on the topic led to the definition of a preprogrammed cycle that would enable locating of external pressure even after full inflation. Locating and relocating of the heel had thus been addressed.

The system also had to be able to detect patient movement to determine when to initiate a relocating cycle. It was agreed that this should also be possible using the pressure sensors. By having the system monitoring the internal pressure in each cell, once the heel has been located and unloaded, changes exceeding a specified threshold can be programmed to trigger a relocating cycle.

Having addressed the main functions, the idea was presented to the project group. Following discussion and reevaluation of all monitoring ideas, it was decided that the pneumatic monitoring system should be the focus of the project. The main reason was its simplicity, and that it required no additional components.

6.2 Scaling of the integrated part.

To determine the size and shape of the integrated part some external

parameters had to be considered, such as the available area of the bed and average patient height. These would determine how much of the surface could be used and where the heels were likely to be. From visiting the bed workshop it was known that modern hospital beds were split into four sections with hinge points connecting them, as previously described. The measurements taken at the meeting revealed that, to avoid being bent

across the hinge point, the integrated part should not exceed 51 centimeters in length.

To ensure the best possible function of the product it was also necessary to find the most likely position of the heel on the bed (lengthways). In order to determine this area the average height of the Swedish population was

retrieved (eLIFE, 2016). Based on the design of modern ward beds it was known that patients would be placed with their hips roughly in the middle of the bed, to enable use of the adjustable back rest etc. Knowing that, the most likely location of the heels could be estimated with the average heights and general proportions of the human body. The general proportions were retrieved from a book on drawing and shows that the hips are at about half

(45)

37 the height of the person. The image referred to is shown in figure 30

(Loomis, 2012).

Figure 30, general proportions of the human figure (Loomis, 2012)

Combining these proportions with the average heights of Swedish males and females an estimation of the average position of the heel could be made.

This was documented as a sketch that is shown in figure 31. Given the fact that the patient’s hips would always be roughly centered on the bed, the general proportions also conclude that all patients taller than one meter would have their heels somewhere in the active area of the mattress.

Figure 31, estimated average position of the patient’s heels.

(46)

38

Determining the size of each cell required consideration of protecting the heel and ankles, while keeping the number of cells to a minimum. Minimizing the number of cells is desirable as it keeps the cost down by reducing the required number of components i.e. valves and sensors. Making the most of the available working area (lowest 50 centimeters) was another factor

influencing the size of each cell.

Specifying the cell size with regard to the protection of the heel and ankles, the size of that area served as reference. With the system determining the location of the heel based on which cells were under external load, all

unloaded cells plus enough additional ones to unload the critical area should be deflated. Figure 32 shows the critical area, measurements were made on a readily available foot and are presumed to be among the higher values expected. The illustrated area does not cover the entire heel, but starts at the estimated starting point of pressure.

Figure 32, area critical to protect.

In the worst case scenario, the heel would trigger a cell with which it is

barely in contact. To ensure that the critical area is protected additional cells must therefore be deflated to the extent that it is completely unloaded. This is illustrated in figure 33 where the heel just barely triggers the cell (3). This leads to the conclusion that an integer number of cells should fit into the critical area to ensure its pressure relief. This coupled with the desire to minimize the amount of cells led to the scale illustrated in figure 33, where one cell is roughly the same size as the critical area. Cells marked with red X:s would be deflated, while the remaining cells under external load would remain inflated. The finalized measure was defined by adapting this to a number that fully utilizes the working area. This resulted in six cells, each 8.5 centimeters wide, which fully utilize the working area of 51 centimeters packed tightly.

(47)

39 Figure 33, the scaled cells.

6.3 Emergency mode

To avoid losing all functionality in case of a power outage it was decided that there should be a valve connecting each cell to the air supply tube, allowing the cells to be inflated or deflated manually if the tubes are disconnected.

Additionally, an emergency switch should be triggered by power loss, making sure that all valves close. This would, in the worst case, result in discomfort if the system happened to be readjusting right at that moment.

The most likely scenario however, is that the system would simply freeze in its current state and not readjust if the patient moves about.

6.4 Integration

Originally the plan was to integrate the product by modifying existing mattresses to accommodating the cells. As the mattress was split into two layers 8 and 6 centimeters thick, this could be done by simply cutting the top layer (6 centimeters) to make room for the cells. While meeting Lasse at the bed workshop however, it became clear that hospitals do not have any possibility of doing such modifications. The group therefore decided that the future product should be an entire mattress and not just the parts necessary for pressure relief. This would most likely be achieved through a future

cooperation with a mattress manufacturer and thus did not affect the work contained within this thesis. In terms of integration this still means the product will be integrated into the ward beds, though not the mattress.

Since mattresses and beds are bought independently of each other previous work on integration still applied.

6.5 Cell material

Given that hygienic aspects, as mentioned in the requirement specification, will affect the cover and not the cells, manufacturing, fire resistance, yield strength and price were the key factors in the choice of material. As the

(48)

40

concept is the deliverable, and thus this selection is done pre-prototype, it should be regarded as a first step proposal and not a final decision.

To find the yield strength threshold a worst case scenario was defined. An assumption was made, that the most stress likely to be put on a cell is if a person was to stand on it. To maintain a worst case scenario, this person is assumed to support his or her entire weight on one foot, placed on a single cell.

Figure 34, estimation of foot pressure area for person on their toes.

Based on the estimations of mass and area shown in figure 34, the resulting cell pressure was calculated as follows.

The strain of the cell was defined using the Young-Laplace equation which is commonly referred to as “ångpanneformlerna” in Sweden (Dahlberg, 2001), stating that stress equals pressure times cylinder inner radius divided by material thickness. Knowing that plastic sheeting can be purchased in

standard thicknesses of up to about 2 millimeters (source), this was used as a limiting value resulting in the minimum yield strength required.

Using CES EduPack relevant materials were screened based on relevant manufacturing processes and fire resistance and then sorted based on their yield strength and price, see the graph in figure 35.

(49)

41 Figure 35, material selection graph

As can be seen in figure X, Polyvinylchloride (PVC) has good margin to the specified yield strength threshold while also being the cheapest alternative.

This makes it the top candidate and the material recommended for prototyping.

6.6 Interface

The group concerned by the interface is the caring staff, as they should be the ones interacting with the product. Their interests are thus what should be considered while designing it. A few basic insights regarding their needs and preferences were gathered from the meeting at the Karolinska

university hospital, as previously mentioned.

The first important aspect was ease of use. Since the caring staff has a multitude of equipment and a general shortage of time, the product should be as close to self-explanatory as possible. The overall goal was of course to minimize the required interaction to the largest possible extent, but some user input was desired.

One desired function was the ability to toggle between two modes, automatic and static, allowing the cells to behave similar to a mattress.

Another function is firmness adjustment, allowing the pressure to be

personalized to give maximum comfort to each patient. Finally there needs to be a reset function to manually initiate a relocation cycle if needed.

Another aspect found to be of interest to the caring staff was practicality i.e.

ability to interact with the product even if wearing gloves or carrying an item in hand. Physical buttons thus seemed preferable over touch screens as they give a more robust impression and could be maneuvered with elbows etc. if required. This was especially true in this case where the limited number of functions require few buttons.

(50)

42

Touch screens would have an advantage from the perspective of another stakeholder, the provider of the product. As it was mentioned in early discussions that additional functions may be added later on, such as more sophisticated movement monitoring for example, the flexibility of a touch screen would enable this without needing to be replaced.

One way of getting some of the flexibility of touch screens while maintaining the physical buttons was using unlabeled buttons that would relate to a screen in such a way that it provided the required information. A sketch illustrating this is shown in figure 36 below.

Figure 36, merger of touch screen and physical buttons.

Further development of this idea resulted in a proposed layout of the interface, using four buttons to control the functions currently available. A fifth button was added to enable future expansion of functions. The resulting idea is shown in figure 37.

Figure 37, idea for interface layout.

(51)

43 The flexibility of the system would allow for various ways of structuring a menu. Some examples were sketched and are shown in figure 38 below.

Figure 38, exemplified alternatives for menu layout.

(52)

44

(53)

45

7 Result

The result of the thesis includes all of the deliverables previously described and each is presented in this chapter.

7.1 Final concept

The final concept is the result of concept two having passed the concept development process. Having been defined to the extent currently possible it is now ready to be turned into the first prototype, which will be used

evaluate the function and further define the design accordingly. At this stage, all of the specified requirements have been met and the final concept thus has the potential to redefine heel ulcer treatment. As was found in the pre study, even small improvements in heel ulcer prevention will have a big impact on healthcare costs and above all, the suffering of patients.

Figure 39 shows the final concept. For illustrational purposes the mattress cover (grey) was cut to reveal the cells, connecting tubes resting on the foam core. The measurements are based on mattresses currently in use at Swedish hospitals, allowing full integration of the final concept into current ward beds. Sterilization requirements are met by the mattress cover, as in products on the market, which eliminates the need for the internal parts to meet it.

Figure 39, the final concept.

When a patient is placed on the bed the pressure added to the cells is detected by sensors located in the controlling unit. The system can then approximate the location of the heel to the last cell showing elevated

pressure. Knowing this, cells are deflated counting from the lower end of the bed until all unloaded cells plus the first loaded one are empty, leaving the heel suspended. Once the heel is relieved of pressure the system will enter monitoring mode, maintaining static inflation levels while monitoring the pressure for movement indication. High enough changes in pressure will trigger a relocating sequence, and the cycle restarts.

(54)

46

The final concept consists of six by two cells stacked closely at the lower end of the mattress. Each cell is connected to the controlling unit via an

individual air tube, resulting in a total of twelve tubes individually controlled by the controlling unit. Each tube connects to the cell via an emergency valve enabling manual inflation or deflation in case of a power outage.

Figure 40 shows the connection between tubes and valves more closely.

Figure 40, close up on the connection between tubes and valves.

Figure 41 and 42 show how the final concept would act when patients of different heights would lie on it. The two dummies featured are about 180 and 160 centimeter tall.

Figure 41, resulting position with taller patient (~180 cm)

(55)

47 Figure 42, resulting position with shorter patient (~160 cm)

The cells will be made from extruded tubes of PVC. They will connect to the controlling unit via emergency valves that allow manual operation, using a manual pump, if the tubes are disconnected. This prevents uncontrolled deflation causing discomfort in case of a power outage. Additionally, the controlling unit will have an emergency trigger that immediately closes all valves in case of power loss. These functions ensure that the worst scenario, in case of power loss, is that the system will no longer be able to reposition itself, causing a maximum of discomfort for the patient. Each section of six cells will be held together by a strip that is either welded or glued to their bottom surface, see figure 43.

Figure 43, joined section of cells.

This connecting strip will have a rough surface providing friction against the foam mattress. This keeps the cells stable during use while allowing easy replacement of a section, should it be damaged. In a later stage this strip should be wider than what is shown in figure X, it is currently shown this way for illustrative purposes.

(56)

48

The unique selling point, or USP, of this concept is mainly the fact that it is the only fully automated preventive measure of heel ulcers on the market.

This eliminates the human error and should thus drastically decrease the risk of developing heel ulcers while hospitalized. On a second note, it is also the first system (identified within the thesis) that uses real time feedback to adapt the surface of the bed to individual patients (current air mattress systems inflate and deflate on a schedule). The flexibility of this concept is another of its strengths. Keeping the mattress as simple as possible, by concentrating all active parts into the controlling unit upgrades and functional changes can be done at any time, while the mattress can be exchanged at the rate of current mattresses. This division also makes the parts sensitive to wear easily replaceable and cheap.

7.2 Brief for controller unit

A summarized version of the brief for the controlling unit will be presented here, consisting of the vital parts which are the requirement specification, the schematic overview and the conceptual cycle of relocating the heel. The full version can be found in appendix 2.

7.2.1 Task and requirements

The work still to be done includes building the controlling unit, programming the required functions and experimenting to set the correct parameters. A prototype of the mattress will be available for experimenting.

Based on the functions defined in the concept the controlling unit must meet the following criteria.

 Enable automatic locating of the heel using pressure input

 Enable control of inflation and deflation of each cell individually

 Enable movement monitoring using pressure input

 Enable relocating of the heel through a reset cycle

 Enable pressure adjustment to compensate for patient weight The deliverables of the thesis should be an assembled prototype of the controlling unit and a program capable of performing the tasks listed in the requirement specification.

(57)

49 7.2.2 Controller unit overview

The basic structure of the controlling unit was defined in the previous thesis work. This structure includes the main category of components needed to perform the required tasks as well as how they connect to each other. The structure is illustrated in figure 44 below.

Figure 44, chart showing the main components of the controlling unit.

As seen in figure 44, the main components of the controlling unit (box) are a pump, electronic valves, electronic pressure sensors and the controlling hardware.

7.2.3 Specific function cycle

Limiting the feedback to use only the internal air pressure was desirable as it keeps the mattress as cheap and simple as possible. This is important since mattresses are replaced regularly as they wear out, making cost a key factor in the success of the product. However, this also complicates the feedback retrieval in some instances. One difficulty with using only internal pressure for feedback was identified and addressed in the thesis. The result is a suggested method, or cycle, to be able to retrieve the input necessary.

During the relocating cycle all cells must be inflated to be able to detect the patient. The problem was that pressure measurements cannot determine whether pressure is caused by the pump, having fully inflated the cell already, or by external load acting on the cell. Therefore, a pressure would lead to the same pressure in all cells, but different stages of inflation, and would prevent locating of the heel. To bypass this problem a specific relocating cycle is suggested.

(58)

50

The first step of relocating the heel would be to fully, and simultaneously, inflate all cells, this requires that the pump is strong enough to do so despite external loads. This would result in all cells being fully inflated and having the same internal pressure, still preventing the heel from being located. To rebalance the pressure all valves would therefore be opened for a short period. This would enable the unloaded cells to quickly depressurize. The loaded cells would also lose pressure, but if the timeframe is calibrated their pressure sensors should still indicate the pressure caused by the patient.

This reasoning is illustrated in figure 45 below.

Figure 45, Expected deflation behavior of loaded and unloaded cells.

As illustrated in figure 45, the deflation properties should differ between loaded and unloaded cells. This would allow the calibration of the resetting time to be calibrated as shown by the green line, resulting in close-to-zero pressure in unloaded cells while loaded cells still detect their external load.

(59)

51 7.2.4 Design proposal and interface

Based on insights gathered at the Karolinska university hospital during an auscultation, a suggestion for the user interface was created. The main

concern was ease of use, which is why current functions are easily controlled via large physical buttons that are easily operated even while holding items or wearing gloves. It was also desired to make it as self-explanatory as possible, since caring staff often work in a stressful environment and may not have time to learn details about individual systems beyond how to operate them. Gathered insights led to a final proposal as shown in figure 46.

Figure 46, design proposal for controller unit and interface.

With current functions there is no need for the menu button and no marking of it is required. However, the button is necessary to enable adding functions at a later time.

(60)

52

Development of an Automatic Heel Ulcer Prevention Mattress

Development of an Automatic Heel Ulcer Prevention Mattress

LINUS GREEN

Development of an Automatic Heel Ulcer Prevention Mattress

Linus Green

Sammanfattning

Abstract

Abbreviation dictionary

Contents

1 Introduction

2 Background

3 Project task

4 Method

5 Implementation

6 Concept development

7 Result