**4. Classification and prioritization of medical equipment for maintenance activities**

Prioritization of medical equipment maintenance should be performed for each new type of device during the inspection received when the device is added to the inventory. The device will then be assigned a test frequency. Subsequently, the maintenance history of the device will be monitored to evaluate the effectiveness of the maintenance program.

The end point of providing an organizational tool to the biomedical or clinical engineer would ensure the safe and efficient performance of medical equipment. The system must be evaluated on criteria such as:


*Operations Management - Emerging Trend in the Digital Era*

temperatures, etc. must be considered.

performances of the system [12].

systems) to complex devices.

tioning of the tested system is not disturbed [13].

emission to determine operating conditions.

are out of danger are emphasized.

method is not so often used in practice.

the maintenance problems and establishing the procedures and the maintenance strategy for equipment must therefore take into account both monitoring and diagnosing at the level of each component, but also the influence of the system variables. Most of the time, the cause of a defect is found in the variations of the process parameters, and a nonintegrative approach to monitoring and diagnosing the system can lead to inefficient actions. Thus, in addition to the most popular techniques of monitoring and diagnosis (vibration monitoring, thermography and tribology), other parameters of a system such as flow rates, voltages, currents,

In systems equipped with computer control or semiautomatic control, most of these parameters are purchased and used in the command and control process. Their type and number vary from system to system, but the algorithm for applying the monitoring and diagnostic procedure is similar. The collection of these parameters, together with the application of the traditional technologies of predictive maintenance, will provide all the necessary data for the analysis of the state and the

Since a large part of the equipment used in the medical field belongs to the category of electromechanical systems, the analysis of the maintenance technologies will focus on these, from the simplest (examples: electric motor-pump type drive

It should be kept in mind that, in any system, the maintenance program will focus on its critical components. A critical component is defined as the element directly involved in the proper functioning of the device, on which the entire system depends, its efficiency and, last but not least, the quality of the product.

Another key parameter that can provide information about one's status of equipment/system is temperature. This is an important indicator of the mechanical, electrical or load conditions applied to a component. Thermography is a predictive maintenance technique that uses instruments that can monitor infrared energy

Infrared scanning is recommended as a regular maintenance procedure in many situations, extracting solid results as quickly as possible and without interrupting process flow, a key benefit to the industry, regardless of the age of the equipment. As an advantage of scanning a large area in a very short time, the ease with which data can be stored and processed for further analysis of images, the high mobility of the thermography camera that can be positioned at any time and place, the thermographic evaluation that is done uninterrupted and equipment inspection staff who

Lubrication fluid analysis can be used to determine mechanical wear, lubrication or fluid condition. The presence of metallic particles in the lubricating fluid suggests the existence of a wear, their analysis providing information on the part subjected to wear. For fluid analysis, it uses complex equipment, which is why this

Some of the technologies for monitoring and diagnosing the state of a system are set out in the following. Vibration analysis is one of the most widely used detection methods to diagnose defects in electromechanical systems. This method measures the vibrations of the system, usually with an accelerometer, and then examines the frequency spectrum generated to identify significant frequencies from the point of view of the state of the equipment. Certain frequencies are typical of the system in normal operation. Changing the amplitude of certain harmonics, for example, can mean the presence of a defect. The data can be collected periodically, using a portable system, or continuously, by installing a continuous monitoring system. A major advantage is that the measurements are fast and noninvasive, and the func-

**238**


The risk assessment was divided into four main areas: clinical function, failure avoidance probability, history of incidents and regulatory or manufacturer requirements. Devices would be evaluated on the aforementioned criteria and be assigned a score. The values would be added and a cumulative score is given for each device type. The total score would act as a quantifiable indicator for the maintenance policy. A total score of 12 or more would indicate a semiannual testing, a score between 9 and 11 would require annual testing, whereas a score of 8 or less would suggest a lesser necessity for annual testing, either biannual or no schedule, depending on clinical use. The end result would be an increase in the cost-effectiveness of the test program, less equipment downtime leading to improved patient care and a higher financial return to direct patient care activities.

To illustrate the applicability of risk assessment criteria, we evaluated two types of devices extensively used in healthcare: the defibrillator and the enteral feeding pump. Defibrillators are devices that correct or prevent arrhythmias (e.g., ventricular fibrillation and ventricular tachycardia) by sending an electrical impulse to the heart. External defibrillators, in particular, send high electrical impulses through the thoracic wall, stopping the independent action of the individual myofibers, so that the intrinsic pacemaker can take over. A set charge, between 0 and 360 J, is generated and delivered through paddles or disposable electrodes through the chest wall to the heart, determining a global contraction. Most defibrillators include an electrocardiograph to monitor the patient's rhythm, while others even include the

pacer function. The clinical use is typically for emergency heart pacing such as severe bradycardia, asystole, pacemaker failure or ventricular fibrillation.

For this particular type of device, the assessment should include electrical safety evaluation—ground wire resistance, chassis and lead leakage—and inspection of parameters' performance, which includes measuring the energy output of the defibrillator throughout its range. This would include determining the value output at the lowest, midlevel and highest settings. The range of error should be with 15% of the set energy level (for 360 J, the output should be ranging from 206 to 414 J). Other performance tests would be determining the output levels at maximum setting for 10 charge cycles. The final output should still be within 15% of the recommended setting and charge time should not exceed 15 seconds. The appraisal for functional assessment frequency would be twice a year (**Table 2**) [15].

Enteral feeding pumps are used in patients who have gastrointestinal complications and who cannot consume adequate nutrients for certain reasons. The feeding solutions are transmitted to the patient through temporary feeding tubes or surgically implanted. These pumps can precisely control the flow of liquid supply solutions that are administered entirely through the digestive tract. These pumps are based on a pump mechanism such as a rotary peristaltic pump, a linear peristaltic pump or a volumetric pump. Most pumps record the dose frequency, dose settings and volume infused into memory. Audible and visual alarms alert the user to flow changes or malfunctions.


**241**

test is annually (**Table 3**).

*Sample risk assessment for enteral feeding pump.*

**5. Maintenance optimization models**

*A Modern Approach for Maintenance Prioritization of Medical Equipment*

**Criteria Risk Score**

Device is used to deliver direct treatment to the patient 4 **4**

Common device failure modes are unpredictable 2 **2**

No history 1 **1**

No requirements 1 **1**

**Total 8 Times per year tested 1 (normal)**

3

No patient contact 1 Device is in contact with the patient who is not critical 2 Device is used for patient diagnosis or direct monitoring 3

Device is used for a life support 5

Maintenance would not impact reliability of the device 1

Specific regulatory requirements dictate preventive maintenance or testing 4

A significant history of incidents exists 2

There are requirements for testing 2

Common device failure is predictable and can be avoided by preventive

**Manufacturers/regulatory requirements for specific schedules**

9 and 11 ml. The measured occlusion pressure must be within 1 psi of the pump occlusion pressure. For an occlusion pressure of 20 psi, the measured pressure must be between 19 and 21 psi. The recommended frequency of the functional

Maintenance costs represent a large part of total cost functioning of health systems. Depending on the specifics of each device, the costs of maintenance can represent from 15 to 60% of the value of the expenses. For the situation in which the equipment works in safe conditions until a certain level of wear or a defect in the initial state has been established, we discuss about preventive and predictive maintenance. In such cases, the equipment will be stopped at an early date, and the repair will only be done where needed. This type of maintenance allows the early detection, localization and identification of the defect or the worn part, as well as the calculation of the operating life in safe conditions of the device. The activity of preventive and predictive type makes possible the planning of the stop, the preparation of the intervention team, the provision of the necessary spare parts and

respectively the minimization of the parking time for repair [16].

Before returning the equipment to medical personnel, it must be ensured that it has been adjusted to the original specific settings. Make sure that the volume of the audible alarms is loud enough to be heard under normal operating conditions [15].

*DOI: http://dx.doi.org/10.5772/intechopen.92706*

**Clinical function**

maintenance

**Table 3.**

**Incident history**

**Problem avoidance probability**

The quantity of volume delivered must be within 10% of the established volume. Thus, for a set volume of 10 ml, the measured volume must be between

#### **Table 2.**

*Sample risk assessment for defibrillator.*

*A Modern Approach for Maintenance Prioritization of Medical Equipment DOI: http://dx.doi.org/10.5772/intechopen.92706*


#### **Table 3.**

*Operations Management - Emerging Trend in the Digital Era*

changes or malfunctions.

**Clinical function**

maintenance

**Incident history**

**Problem avoidance probability**

*Sample risk assessment for defibrillator.*

pacer function. The clinical use is typically for emergency heart pacing such as severe bradycardia, asystole, pacemaker failure or ventricular fibrillation.

evaluation—ground wire resistance, chassis and lead leakage—and inspection of parameters' performance, which includes measuring the energy output of the defibrillator throughout its range. This would include determining the value output at the lowest, midlevel and highest settings. The range of error should be with 15% of the set energy level (for 360 J, the output should be ranging from 206 to 414 J). Other performance tests would be determining the output levels at maximum setting for 10 charge cycles. The final output should still be within 15% of the recommended setting and charge time should not exceed 15 seconds. The appraisal for

functional assessment frequency would be twice a year (**Table 2**) [15].

For this particular type of device, the assessment should include electrical safety

Enteral feeding pumps are used in patients who have gastrointestinal complications and who cannot consume adequate nutrients for certain reasons. The feeding solutions are transmitted to the patient through temporary feeding tubes or surgically implanted. These pumps can precisely control the flow of liquid supply solutions that are administered entirely through the digestive tract. These pumps are based on a pump mechanism such as a rotary peristaltic pump, a linear peristaltic pump or a volumetric pump. Most pumps record the dose frequency, dose settings and volume infused into memory. Audible and visual alarms alert the user to flow

The quantity of volume delivered must be within 10% of the established volume. Thus, for a set volume of 10 ml, the measured volume must be between

No patient contact 1 Device may make contact with the patient who is noncritical 2 Device is used for patient diagnosis or direct monitoring 3 Device is used to deliver direct treatment to the patient 4

Maintenance would not impact reliability of the device 1 Common device failure modes are unpredictable 2

No history 1

No requirements 1

Common device failure is predictable and can be avoided by preventive

**Manufacturers/regulatory requirements for specific schedules**

**Criteria Risk Score**

Device is used for a life support 5 **5**

Specific regulatory requirements dictate preventive maintenance or testing 4 **4**

A significant history of incidents exists 2 **2**

There are requirements for testing 2 **2 Total 13 Times per year tested 2 (hight level)**

3

**240**

**Table 2.**

*Sample risk assessment for enteral feeding pump.*

9 and 11 ml. The measured occlusion pressure must be within 1 psi of the pump occlusion pressure. For an occlusion pressure of 20 psi, the measured pressure must be between 19 and 21 psi. The recommended frequency of the functional test is annually (**Table 3**).

Before returning the equipment to medical personnel, it must be ensured that it has been adjusted to the original specific settings. Make sure that the volume of the audible alarms is loud enough to be heard under normal operating conditions [15].

## **5. Maintenance optimization models**

Maintenance costs represent a large part of total cost functioning of health systems. Depending on the specifics of each device, the costs of maintenance can represent from 15 to 60% of the value of the expenses. For the situation in which the equipment works in safe conditions until a certain level of wear or a defect in the initial state has been established, we discuss about preventive and predictive maintenance. In such cases, the equipment will be stopped at an early date, and the repair will only be done where needed. This type of maintenance allows the early detection, localization and identification of the defect or the worn part, as well as the calculation of the operating life in safe conditions of the device. The activity of preventive and predictive type makes possible the planning of the stop, the preparation of the intervention team, the provision of the necessary spare parts and respectively the minimization of the parking time for repair [16].

Predictive maintenance represents a superior qualitative leap in a modern maintenance system, regardless of the domain or the specific production, because it offers all the information needed for the following:


The common premise from which the predictive maintenance starts is that the periodic or continuous monitoring of the mechanical, electrical or other indicators of the functioning of the systems or processes can provide the data necessary to ensure the maximum interval between the repair and maintenance works, respectively, to minimize the cost of interruptions of maintenance. Unplanned maintenance can be the cause of possible failures, sometimes major. However, predictive maintenance is more than that. It is in fact the means of improving and increasing the productivity, product quality and overall efficiency of the systems in question. Predictive maintenance is actually a philosophy or attitude that, based on operating conditions, allows the optimization of the entire medical system. A comprehensive management of predictive maintenance uses the best methods to obtain the operating parameters of the component subsystems of a medical system, on the basis of which it will schedule maintenance and repair activities. Including predictive maintenance in the general maintenance program optimizes the availability of devices and equipment and greatly reduces maintenance costs. By using the records of the entire care of historical repair components and maintained maintenance, we can make a mathematical prediction model for the entire world.

Classifications of different types of failures and the establishment of policies for analysis involve three different levels: system level, failure peak and component level. Results analyzed can be set for a model for optimizing maintenance/ inspection.

#### **5.1 Rejects detection model**

It is considered a continuous process so that it can be put into operation or rejected (scrap). The way of monitoring the functionality is as follows: first, check each product; continue checking until the consecutive *k* linear products are reached (full inspection). From this point, the inspection of the equipment is no longer deterministic, "piece by piece"; they will be chosen randomly, independently of the other, with probability α. Continue random monitoring (partially verified) until a defect is discovered, and then revert to previous monitoring and so on. Suppose the probability of a product being defective is *q*. It is understood that if a problem is found, the item is removed temporarily or permanently. \_

$$\frac{\left(\frac{1}{p}\right)^k - 1}{q}, \text{ where } p = 1 - q \tag{1}$$

Average of all relevant product cycle is equal to:

$$\text{It cycle is equal to:}$$

$$P\_1 = \frac{\left(\frac{1}{p}\right)^k - 1}{q} + \frac{1}{aq} \tag{2}$$

**243**

staff.

equipment.

management.

*A Modern Approach for Maintenance Prioritization of Medical Equipment*

C2 amount, corresponding to the provision of the equipment.

*∫*0

the level of finding and solving the cause of the defect itself.

themes according to current standards (**Figure 3**) [17–20].

repairs that involve certain assumptions and challenges.

**6. Life cycle of medical equipment**

Some assumptions are as follows:

The average length of a life cycle of equipment can be expressed as:

Depending on the distribution of H, which is usually uniform (0, T0), where T0 is a standard period, depending on the case, for example 10 years and costs C1 and C2, one can estimate the value of T, which will reduce to a minimum the cost of having an older, optimizing device. As in the field of health care, failure prevention is more effective than focusing on remedying them. Repairs are almost always expensive, requiring overspecialized personnel and often expensive parts. However, corrective maintenance is a permanent component of medical technology

Corrective maintenance allows a device to maintain its full performance of functions, through effective interventions at the time of a problem. However, this action must be well planned, because it acts not only on the level of symptoms, but also on

Users and technical staff have the obligation to maintain medical equipment at a level of safety as high as possible, compared to other types of usual equipment. Most complex medical equipment works, for example, in the intensive care unit. They have an electrical connection that in certain situations of first defect can create injuries or even death of the patient by electric shock. Patients connected to such medical equipment are not able to respond to dangerous conditions or pain. Other types of medical equipment work to support life, and a problem, sometimes even minor in some respects, can lead to the death of the patient when the equipment is used incorrectly or is poorly maintained. The life cycle of medical equipment, from the point of view of media technology management, comprises 4 stages and 9

An important stage in the life of medical equipment is that of maintenance and

• Maintenance culture exists and is respected by the technicians, users and other

• Technical staff are present, trained and know how to maintain and repair the

It can be shown that the proportion of undetected defective items is given by: *P*2 = \_ *q*(1 − *P*1) (1 − *qP*1)

Another model that offers good results when used in this field is known as "replacing a durable good." This is based on the assumption, for example, that the service life of the equipment is represented by a continuous random variable with the distribution function H and the density h and that a policy to replace the good says that it will happen if it has a major failure or if it is still in operation, it is acceptable to reach a certain "age," say the T years. We assume that the price of similar new equipment is C1, and when the equipment fails, we seriously consider a

(3)

*<sup>T</sup> xh*(*x*)*dx* + *T*[1 − *H*(*T*)] (4)

*DOI: http://dx.doi.org/10.5772/intechopen.92706*

*A Modern Approach for Maintenance Prioritization of Medical Equipment DOI: http://dx.doi.org/10.5772/intechopen.92706*

*Operations Management - Emerging Trend in the Digital Era*

offers all the information needed for the following:

• Early detection of the defects

• Location

• Diagnosis of defects

Predictive maintenance represents a superior qualitative leap in a modern maintenance system, regardless of the domain or the specific production, because it

• Calculation of the operating life in safe conditions of the medical equipment

The common premise from which the predictive maintenance starts is that the periodic or continuous monitoring of the mechanical, electrical or other indicators of the functioning of the systems or processes can provide the data necessary to ensure the maximum interval between the repair and maintenance works, respectively, to minimize the cost of interruptions of maintenance. Unplanned maintenance can be the cause of possible failures, sometimes major. However, predictive maintenance is more than that. It is in fact the means of improving and increasing the productivity, product quality and overall efficiency of the systems in question. Predictive maintenance is actually a philosophy or attitude that, based on operating conditions, allows the optimization of the entire medical system. A comprehensive management of predictive maintenance uses the best methods to obtain the operating parameters of the component subsystems of a medical system, on the basis of which it will schedule maintenance and repair activities. Including predictive maintenance in the general maintenance program optimizes the availability of devices and equipment and greatly reduces maintenance costs. By using the records of the entire care of historical repair components and maintained maintenance, we

Classifications of different types of failures and the establishment of policies for analysis involve three different levels: system level, failure peak and component level. Results analyzed can be set for a model for optimizing maintenance/

It is considered a continuous process so that it can be put into operation or rejected (scrap). The way of monitoring the functionality is as follows: first, check each product; continue checking until the consecutive *k* linear products are reached (full inspection). From this point, the inspection of the equipment is no longer deterministic, "piece by piece"; they will be chosen randomly, independently of the other, with probability α. Continue random monitoring (partially verified) until a defect is discovered, and then revert to previous monitoring and so on. Suppose the probability of a product being defective is *q*. It is understood that if a problem is

*<sup>q</sup>* , where *p* = 1 − *q* (1)

*<sup>q</sup>* (2)

\_1

can make a mathematical prediction model for the entire world.

found, the item is removed temporarily or permanently.

Average of all relevant product cycle is equal to:

( \_1 *p*) *k* \_ − 1

*<sup>P</sup>*1 = ( \_1 *p*) *k* \_ − 1 *<sup>q</sup>* +

**242**

inspection.

**5.1 Rejects detection model**

It can be shown that the proportion of undetected defective items is given by:

$$\text{ign206}$$

$$\text{ion of undirected defective items is given by:}$$

$$P\_2 = \frac{q\{1 - P\_1\}}{\{1 - qP\_1\}}\tag{3}$$

Another model that offers good results when used in this field is known as "replacing a durable good." This is based on the assumption, for example, that the service life of the equipment is represented by a continuous random variable with the distribution function H and the density h and that a policy to replace the good says that it will happen if it has a major failure or if it is still in operation, it is acceptable to reach a certain "age," say the T years. We assume that the price of similar new equipment is C1, and when the equipment fails, we seriously consider a C2 amount, corresponding to the provision of the equipment.

The average length of a life cycle of equipment can be expressed as:

$$\left[\right]\_0^T \ge h(\infty)d\infty + T\left[1 - H(T)\right] \tag{4}$$

Depending on the distribution of H, which is usually uniform (0, T0), where T0 is a standard period, depending on the case, for example 10 years and costs C1 and C2, one can estimate the value of T, which will reduce to a minimum the cost of having an older, optimizing device. As in the field of health care, failure prevention is more effective than focusing on remedying them. Repairs are almost always expensive, requiring overspecialized personnel and often expensive parts. However, corrective maintenance is a permanent component of medical technology management.

Corrective maintenance allows a device to maintain its full performance of functions, through effective interventions at the time of a problem. However, this action must be well planned, because it acts not only on the level of symptoms, but also on the level of finding and solving the cause of the defect itself.
