Precision and reliability study of hospital infusion pumps: a systematic review

Abstract

Background

Infusion Pumps (IP) are medical devices that were developed in the 1960s and generate fluid flow at pressures higher than that of normal blood pressure. Various hospital sectors make use of them, and they have become indispensable in therapies requiring continuity and precision in the administration of medication and/or food. As they are classified Class III (high risk) equipment, their maintenance is crucial for proper performance of the device, as well as patient and operator safety. The principal consideration of the pump is the volume infused, and the device demands great attention to detail when being calibrated. A lack of necessary care with this equipment can lead to uncertainty in volume and precision during the administration of substances. Because of this, it is essential to evaluate its reliability, to prevent possible failures at time of execution. This control aims at the quality of the intended infusion result, becoming an indication of quality.

Methods

This systematic review summarizes studies done over the last 10 years (2011 to December 2021) that address the reliability and accuracy of hospital infusion pumps, in order to identify planning of maintenance and/or other techniques used in management of the equipment. The Prisma method was applied and the databases utilized were Embase, MEDLINE/Pubmed, Web of Science, Scopus, IEEE Xplore, and Science Direct. In addition, similar reviews were studied in Prospero and the Cochrane Library. For data analysis, softwares such as Mendeley, Excel, RStudio, and VOSviewer were used, and Robvis helped in plotting risk of bias results for studies performed with Cochrane tools.

Results

The six databases selected produced 824 studies. After applying eligibility criteria (inclusion and exclusion), removing duplicates, and applying filters 1 and 2, 15 studies were included in the present review. It was found that the most relevant sources came from the Institute of Electrical and Electronics Engineers (IEEE) and that the most relevant keywords revolved around the terms (“device failure”, “infusion pumps”, “adverse effects”, “complications”, etc.). These results made clear that there remains substantial room for improvement as it relates to the study of accuracy and reliability of infusion.

Conclusions

We verified that the reliability and precision analysis of hospital infusion pumps need to be performed in a more detailed and consistent way. New developments, considering the model and IP specification, are intended, clearly explaining the adopted methodology.

Background

The first uses of intravenous therapy were recorded in the 1600 s, and were done with experiments using animal feathers, and bladders as infusion materials [1]. Following the Second World War, this type of therapy was introduced as a routine practice in healthcare [1]. The first automatic equipment, built by the German scientist Watkins, had electromechanical characteristics; and was called the “chronofuser” [2]. At that time, knowledge of micro-controlled devices and digital electronics was almost non-existent [2]. A variety of medical treatments are performed through infusion, including use of sedatives, analgesics, hormones, parenteral nutrition, and chemotherapeutic agents, among others. Infusions may have varying degrees of duration, depending on their specific use, and require great safety and precision [3]. With the modernization of health systems, Infusion Pump (IP) equipment was developed and implemented in healthcare centers in the 1960 s [4]. It is a device capable of generating the flow of a given fluid, at pressures that are higher than that of regular blood pressure for the intended location [5].

We found a small number of published studies describing the evaluation of reliability in infusion pumps. In view of this, more research related to the types of procedures performed and their related maintenance programs is necessary to ensure stable operation. These technological analyses are necessary to prevent sudden interruptions, as well as to minimize worst case scenarios, such as complete machine shutdown. As such it is essential to verify which types of testing and evaluation procedures have been conducted to keep them working properly.

To the best of our knowledge, a systematic review on the subject has not yet been published. Therefore, considering the importance of studies on equipment reliability (errors related to installation, maintenance, failures, and operating conditions), the objective of this review is to summarize published data on dependability and accuracy in infusion pumps. We investigated whether these and other variables have been evaluated and/or quantified over the last 10 years, as well as looking at current evidence for potential strategies to minimize risk of failure for this critical medical equipment.

Infusion pumps

The use of IPs, because of their specific alarms and controls, allows for accurate and safe infusion of liquids into the body [4]. The literature states that they require positive pressure to function which requires a directing mechanism. These operating principles can be peristaltic (rotary or linear), syringe or piston driven [6]. A modern infusion system consists of a percutaneous instrument (intravascular catheter) that passes through the skin to infuse fluids into a vein, a tube for transporting the liquid or a reservoir, containing the fluid to be delivered, which can be a bag or syringe, and a flow control device capable of stopping or regulating the infusion [3]. This system is shown in Fig. 1, including the IP.

Fig. 1

Representation of an infusion system including the infusion pump

As it is a mechanism directly connected to the patient, the attention to detail and care used with this equipment is critical. Intravenous infusions are the most common type of infusion and if they fail can cause aggravating problems for the patient and/or worsen their health condition, for example, causing venous spasm, pulmonary edema, and phlebitis [7, 8].

In the hospital environment, infusion pumps are most commonly found in Intensive Care Units (ICUs), Neonatal Intensive Care Units (NICU), and Surgical and Oncology Centers [9]. The process performed by infusion pumps is the automation of a repetitive hospital technique and is of great relevance to healthcare professionals, and directly impacts patients’ lives. Despite their importance, they are medical devices associated with the highest number of technical complaints, or adverse events, according to the techno-surveillance unit of the Agência Nacional de Vigilância Sanitária (ANVISA) [10]. This device is one of the most used Medical Assistance Equipment’s in ICU’s and is responsible for 19.4% of all adverse effects originated in the general context of the hospital environment, resulting from failures in drug administration [11].

The COVID-19 pandemic led to a dramatic increase in the use of ventilators. This increase, and the accompanying discomfort experienced by patients on breathing machines resulted in greater demand for the delivery of sedative infusions with bags requiring changing every 2–3 h [12]. Faced with this challenge, in April 2020 the Food and Drug Administration published changes related to the handling of pumps. The new guidelines particularly involved the ability to remotely monitor and adjust pump parameters throughout this public health emergency [12].

These decisions increased availability of equipment and helped reduce exposure of healthcare professionals to patients afflicted by the disease [13]. Other studies addressed the use of the devices outside patient beds and analyzed the handling and safety of the user [13,14,15]. However, despite this commitment to minimizing impacts of the pandemic, Nova Scotia Health’s leadership noted that a shortfall of infusion pump availability still persists in hospital settings [16].

Reliability in medical equipment

Reliability of medical instruments can be understood as the probability of performing a given task without failure during a specific time interval under stated conditions and it is increasingly important for patient security. New developments in this area permit the use of new instrument models for in-home treatment of illnesses, as well as those for traditional hospitals and clinics settings. Hence, reduction and prevention of operating failures is critical and can be achieved by applying reliable regulation, rigorous maintenance and a warranty control plan [17].

Governments play an important role in the dependability of medical products. In countries like the United States and the European Union, regulations are established to guide direction of design and development phases of medical instruments, where manufacturers must anticipate and address any and all deficiencies that could lead to potential failure. In this respect, there are three distinct classes of instruments that require different approaches: general controls (class I), general and special controls (class II) and general controls and market approval (class III). According to Hedge (2008), infusion pumps are heavily regulated items and require specific labeling, adherence to performance standards and product tracking. As a result, the Medical industry must have solid programs in place that address concept, design, prototype and manufacturing phases as part of any new product development. Products already in use can be assessed using dependability tests, performed on units to induce failure modes, or to anticipate potential failures and implement corrective actions [17]. The author highlights known reliability tools that are recommended at each step, but claims that they are just starting to be implemented [17, 18].

Infusion pumps are instruments largely used in hospitals and their consistent, quality operation must be guaranteed. Failures in operation are generally registered in databases, together with maintenance data and classified as failure modes, operation times, repair times, and actions performed. In general, manufacturers are called to make repairs when necessary. However, due to the sheer number of infusion pumps in use at any given time in the hospitals reliability tests, and correlated data science tools could be applied to extract valuable information. The regular and consistent use of these types of equipment requires the utmost care to ensure safety of patients and operators. It is recommended that multiple studies on the reliability of medical equipment should be carried out around the world, to reduce possible errors in diagnoses, tragedies, injuries, economic losses, and other possible damage [18, 19]. Our proposal begins to close gaps in existing research.

Methods

The present systematic review was conducted to summarize the last 10 years of research on the reliability and accuracy of hospital infusion pumps. The identification of effective maintenance planning and other equipment management techniques is the goal of this research.

Protocol and registration

The present study was conducted according to PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines [20]. The protocol for this study was registered with the International Prospective Register of Systematic Reviews (PROSPERO) [21] under registration number: CRD42022304368. We used reference management software Mendeley^® and classified the folders according to the databases consulted. We used the RStudio (Bibliometrix library), and VOSviewer 1.6.8. software.

Eligibility criteria

Inclusion Criteria

This review established inclusion criteria using the PICO approach (Population, Interest, Comparison, Outcome) [21]. (P) Research problem/Patient, population: hospital infusion pumps; (I) Interest: trials, tests, calibration; (C); not applicable; (O) Primary outcome: reliability study. Secondary outcome: main causes of pump failures, and precision rates. Filters were applied to limit the period (2011–2021), and to select conference (IEEE) and journal articles as well as language (English or Portuguese).

Exclusion Criteria

The following types of studies were excluded: (i) reviews, letters, personal opinions, book chapters and conference abstracts; (ii) articles that did not meet the inclusion criteria; (iii) studies of other types of equipment, such as implantable pumps; (iv) publication in languages other than English or Portuguese; (v) studies where the full paper copy was not available; (vi) low quality studies and (vii) studies done outside of the selected period (2011–2021).

Information sources and search strategy

The creation of one single type of search strategy is generally not possible. Therefore, we elaborated on different combinations using Boolean operators, Individual search strategies for each of the following bibliographic databases were looked at: Embase, MEDLINE/PubMed, Web of Science, Scopus, IEEE Xplore, and Science Direct and the strings adapted for each database are listed in Table 1. In addition to the 6 databases, we verified the existence of similar studies in the Prospero and Cochrane Library, where we did not obtain results. The database search was conducted on October 10, 2021, with no time restriction. Duplicate references were removed using Reference Manager Software (Mendeley^®). Articles were organized and analyzed systematically as the review provides an emerging observance in the COVID-19 pandemic, a fact that intensified the use of this equipment. To produce literature that was cohesive and integrated properly, we defined time limitations as contributions from January 2011 to December 2021, a period recognized for significant updates in the area of medical equipment enhanced standard adoption, for example, NBR IEC 60601–2-24.

Table 1 Database search strings

Study and selection

The articles were selected in two phases: reading of titles and abstracts (phase 1) and reading of full text (phase 2). In phase 1, six authors (Silva, M. S.; Araújo, J. L.; Nunes, G. M.; Rosa, S. S. R. F.; Rosa, M. F. F.; Piratelli-Filho, A.) in blind pairs reviewed titles and abstracts of all references identified by the electronic databases which appeared to meet inclusion criteria and that respected eligibility criteria.

During phase 2, six pairs of authors (Silva, M. S.; Araújo, J. L.; Nunes, G. M.; Rosa, S. S. R. F.; Rosa, M. F. F.; Piratelli-Filho, A.) independently performed full text reviews of the articles, applying exclusion criteria. A third author was consulted in cases of disagreement (Luz, G. V. S.). Finally, the diagramming and extraction of the data obtained was organized in 3 written by the authors of the study and following verification protocol of PRISMA 2009 [20]. The excluded studies and the reason for exclusion are presented in the Additional file 1.

Analysis of the findings

After the search and selection of studies of interest, the extracted data was tabulated using Excel spreadsheets for the development of analyses. To archive the bibliographic data, we utilized Mendeley software, from which we exported information to be evaluated with the help of the RStudio software (Bibliometrix library) and also VOSviewer 1.6.8. These tools make it possible to explore data such as scientific journals, identifying those with the greatest impact, the most influential authors, most common and relevant keywords, and other analyses.

Risks of bias and quality in individual studies

Due to the heterogeneity of the techniques presented, we used narrative description with associated information, classification of the quality of studies, and minimization of bias through the Cochrane Collaboration Tool for Bias Risk Assessment applying a “Generic” platform. The platform can be adapted and has 7 domains, namely: (1) random sequence generation, (2)allocation concealment, (3) blinding of participants and personnel, (4) blinding of outcome assessment, (5) incomplete outcome data, (6) selective reporting and (7) other sources of bias. Hence, we evaluated the quality of the studies, where 4 independent reviewers (Silva, M. S., Rosa, S. S. R. F., Nunes, G. A., Piratelli-Filho, A.) analyzed risk of bias applying the methods described above. To visualize risks, we plotted figures using the Robvis tool, which is a web application aiding in the visualization of risk-of-bias [22]. Articles were classified as “high”, “low”, “no information” and “unclear” in regards to the risk of bias and “none” for missing data.

Results

Selection of studies

The search process for studies in 6 databases resulted in 824 studies. We applied criteria of inclusion, exclusion, and removal of duplicates. After employing these parameters we obtained the Prism diagram shown in Fig. 2.

Fig. 2

PRISMA 2020 diagram: methods and filters adopted in bibliographic-searches and selection criteria according to PRISMA 2020 protocol [20]

Process of analysis of the findings

After removing duplicates, 765 studies remained to be analyzed according to the first filter, that is, titles and abstracts. The main journals where these articles were published are listed in Table 2, which records prevalence of the engineering area. We note that the most relevant sources on the intended topic are predominantly found in the databases of the Institute of Electrical and Electronics Engineers (IEEE). For searches in the IEEE, we considered conferences and other related events, as they generally have high quality standards for acceptance. Of the 15 sources listed in Table 2, 8 are from the IEEE bases, the others are varied and from different health areas.

Table 2 Main journals that concentrate articles on the topic

The IEEE databases, in addition to leading in the number of studies in the area, have growth in publications from 2011 to 2021, as shown in Fig. 3. The journal IEEE Transactions on Biomedical Engineering presents 06 studies (year 2011), 07 studies (year 2012), 11 studies (in 2013), 15 studies (in 2014), 17 studies (year 2015), 20 studies (2016), 21 studies (2017), 22 studies (years 2018 and 2019), 27 (in 2020) and 30 publications on the subject in the last year (2021), as shown in blue in the graph. Then, we can see the Journal Blood, which remained constant from 2019 to 2021, with 17 studies each year, and the Journal of Diabetes Science and Technology with the same frequency of publications as the previous one (2019–2021), but with 14 studies per year.

Fig. 3

Representation of publications in the area from 2011 to 2021

Using VOSviewer software, we analyzed the keyword network of search findings, as shown in Fig. 4. The nodes and their respective sizes represent the number of times, proportionally, that each word was cited. These nodes are connected if the words are co-quoted, in other words, cited by the same article. The connection between two nodes intensified depending on the number of co-citations. As a parameter, the software was informed that, to be considered a co-citation each word had to be mentioned at least 5 times, resulting in 205 nodes. This analysis was performed before the inclusion phase of the studies, allowing us to observe the strength of the words that involve insulin pumps and implantable ones, such as “insulin”, “sugar” and other terms. For the inclusion phase, we did not consider insulin pumps, so the area of interest of this review revolved around terms predominantly in highlighted in blue (“device failure”, “infusion pumps”, etc.) and green (“adverse effects”, “ complications”, etc).

Fig. 4

Bibliometric analysis of the 765 works obtained by searches in the data base. For this, the VOSviewer program version 1.6.1 was used, with the configuration: “full counting”, with at least 5x of terms co-occurrence

Following the analyses presented above, with all the search findings (n= 765), we considered studies included after filters 1 and 2 were applied (n= 15). Using RStudio software, we adopted the Bibliometrix library to evaluate titles of the selected studies and plotted the word cloud shown in Fig. 5. The cloud represents the 70 most frequently used words in titles, with the term “pump” appearing most often (8 times) among the 15 titles.

Fig. 5

Word cloud produced from the words that appear most in the titles

From the included studies, 3 list Elsa Batista as the primary author [23,24,25], highlighting her as being influential in this area of studies, as shown in Fig. 6. João A. Sousa is the second most cited author in the sample.

Fig. 6

Most relevant authors among the inserted studies

The included studies and their main characteristics are presented in Table 3. The 15 studies included were developed in different countries with 20% of the studies done in Portugal, 13% in Brazil, 13% in the United States, and 7%, respectively, for Canada, France, Hong Kong, Iran, Korea, London, Spain and Switzerland.

Table 3 Characteristics of studies included in the review

Quality and risk of bias individual studies

In evaluating the introduced studies, the Robvis “Generic” tool [22] showed that two studies had a high risk of bias [28, 29]. The domains in which they have problems are associated with 4) blinding of outcome assessment, 5) incomplete outcome data and 7) other sources of bias. We also saw that two other works had some associated problems and were not clear about the overall risk [30, 37], as shown in Fig. 7. In terms of criteria, we used “high” for high risk of bias, “low” for low, “no information” for missing data, and “unclear” when the information provided was incomplete or difficult to understand.

Of the included articles, most present a well-defined study and low risk of bias (n= 11). Accordingly, we observed that 14 studies discussed in this systematic review are of high quality for domain 1 and 13 studies for domain 6, both being above 75%, thus adding considerable quality to the included studies.

Fig. 7

Quality analysis of studies plotted on Robvis using the “Generic” tool. Of the included studies, 11 had a low risk of bias

Outcomes of the studies

Following quality analysis of the studies, we developed a table that summarizes results of the research outcomes included in this review. In Table 4, we check the studies, the intervention adopted in each one, and their respective outcomes.

Table 4 Outcomes of the inserted studies

Discussion

Data and infusion pumps analyzed

This was a systematic review of 15 studies that evaluated the condition of infusion pumps. The review investigated whether accuracy and reliability had been analyzed during the last 10 years. We verified that to proceed with these types of analysis, it is not always necessary to have the equipment or to know specifically the manufacturer/model. In the studies [26, 29, 30, 34 and 37] the data considered referred to reports, databases, hospital files, recall data, and other files capable of supporting information for the evaluations. Based on this information, the studies followed different lines to investigate operation of the pumps.

Of the included studies, 10 had specific IPs to evaluate and explore in general. Study [23] tested syringe and peristaltic pumps to investigate different flow methods, occlusion phenomena, and other parameters, to improve infusion accuracy. Syringe IPs are used with disposable syringes and have a complex system of hardware and software designed to monitor the infusion process, probability of infused liquid errors, pump information and other parameters [38]. In this pump model, the syringe acts as a reservoir for the fluid and most of the time its capacity is 60 ml or less. The syringe plunger moves from the pump controls that push fluid and regulate its flow [3].

Studies [24, 25, 32, 33, 35 and 36] also considered different aspects of syringe pumps. Study [24] applied a gravimetric method to assess flow rate of syringe pumps under different conditions and in their study the following year used interferometry to compare with the gravimetric method [25]. Another study that addressed syringe BI associated with the dynamic gravimetric method was proposed by [33]. In it, the pumps were evaluated and error and uncertainty were obtained through the flow. It was concluded that pulsating flows and rates must be measured as they interfere with pump operation. The technique of evaluating BI using gravimetry was identified in 3 studies.

In study [35] infusion accuracy was also investigated for infusion accuracy. However, they were subjected to tests in different environments with variations in height and density of the substance, to determine if such factors would influence precision. For this purpose, pumps were placed at the distal infusion outlet level and later moved up and down. In addition, using syringe pumps, study [36] performed vertical displacement maneuvers and occlusion of the infusion line, to conclude that the problems that affected the tested sets can be partially solved with incremental improvements to the equipment. As well as in studies [35 and 36]. In study [32] the impact of height variation on the flow rate during the use of syringe pumps is also evaluated, reaching similar conclusions. Thus, we can see that the positioning of the equipment can interfere with its operation and has been considered in the literature.

In addition to the use of syringe pumps, we identified error analysis adopting peristaltic BI. The peristalsis propulsion model is the most commonly used, as it allows for a constant flow of substances without damage to liquids and its flow will depend on the thickness of connected tubes. The main limitation of peristaltic pumps is the high cost of the equipment [39]. Of the listed studies [28 and 31] evaluated pumps of this category. The analysis of [28] discusses data entry errors in digital displays. The data refer to registration errors when typing prescriptions using the 5-key keyboard and the results were used to determine probability distribution for the errors found. Study [31] also used a peristaltic pump to carry out a situational diagnosis with more than 300 pieces of equipment. This approach focused on preventive maintenance through descriptive and quantitative research. The diagnosis revealed a lag and outdated preventive maintenance for a technology park at a federal hospital in Rio de Janeiro, Brazil. Such findings present evidence of a lack of precision and reliability in the infusions performed in the environment, where the study was carried out and further reinforce the need for periodic maintenance.

Despite adopting a quantitative measure of 50 IPs, study [27] does not identify the characteristics of the evaluated equipment. In this study, tests were performed to examine if the flow rate and volume are compatible with the schedule and if the alarms are working properly. In addition, they consider fundamental aspects of being assessed, which relates to the guarantee of patient and operator safety. It is noted that the equipment model is a factor to be considered in studies aimed at precision and reliability, as well as the position of use.

Analysis of precision and reliability of studies

Medical products must go through the following phases to achieve reliability: analysis, testing, associated areas, reliability measures, and failure data [40]. Fogliatto et al. details that a reliability program aggregates the institution of procedures and routines to manage dependability in four phases of a product’s life: design and development, manufacture and installation, operation and maintenance, and disposal [41]. Thus, in this study, we specifically evaluated the operation and maintenance phase.

Among the selected studies, [26, 30, 31 and 34 do not have direct interventions for accuracy or reliability analyses. However, some variables are correlated with safety of infusions. The study [26] evaluates pump-related errors and divides them into (a) technology-related errors and (b) non-technology-related errors. In (a) there was a division between sociotechnical errors (which have human interaction) and device errors (related to the technical defects of the device). Equipment errors and failures must be extensively investigated and discussed, as they are obstacles to achieving accuracy and reliability. Human errors can be minimized with qualified labor and training, while technical errors require high rigor in the product design and development stage.

In study [34] incident reports involving IP were evaluated using Failure Modes and Effects Analysis (FMEA) techniques. This technique made it possible to detect risk points in the use of pumps and the variables associated with failures, which are essential in the investigation and comprise the methods for measuring reliability [41]. On the other hand, considerations in study [30] were made from IP recalls, which examined the number and type of failures in various pumps. The study also addresses factors of administration errors and usability, arguing that innovation and technology development should be associated using the human factors approach.

Finally, study [31] did a situational diagnostic search related to preventive maintenance of the pumps. Through descriptive and quantitative research, they found that less than 10% of the pumps’ technological parks were up to date in terms of maintenance, 54.5% had expired preventive maintenance records and 29.9% had no record of maintenance at all. The results and conclusions of study [31] illustrate negligence with equipment that can cause damage to health, in a basic principle that is preventive maintenance. Therefore, the reliability and accuracy of the equipment is not guaranteed for either the operator or the patient.

Study [42] notes that precision can be defined as the degree of conformity between a series of observations of the same random variable, and that the spread of the probability distribution is an indicator of precision. In this context, study [23], analyzing syringe infusion pumps, investigated rapidly changing flow rates, liquid mixing behavior, and occlusion phenomena in multi-infusion systems, aiming to improve the accuracy of delivered dosage, particularly for flows as low as 100 nL/min. The study concludes that by improving the accuracy of flow rate measurement in drug delivery devices, with the development of new measurement methods, dosing errors can be reduced. This can be achieved by a broad uptake of traceable infusion calibrations and enhanced knowledge of calibration for drug delivery devices in clinical settings, especially in the case of multiple infusion systems.

Similarly as in the previous one, also focused on flow, study [27] performed tests to verify flow rate, volume, and bolus delivered by the device at a required accuracy rate. They verified the safety of the device for the patient and operator, as well as confirming that occlusion alarms are activated during emergency conditions. The study also investigated the electrical safety of infusion pumps.

Another study that approaches flow as a theme is number [24]s. Study [24] investigates the influence of rapidly changing flow rates in general with the gravimetric method, due to a predefined flow rate change. In addition, a multi-infusion setup is developed to analyze flow rates of fluids and their compositions at the exit of the infusion line. They concluded that dosing errors can be reduced by improving flow rate measurement of devices and that this can be achieved by capturing traceable low-flow and ultra-low-flow infusion calibrations, in addition to a better understanding of dosage administration calibration in clinical environments. As in the previous study, [33] adopted the gravimetric method to analyze the IP. The authors conclude that pulsating flows and rates can be measured as they interfere with pump operation. They confirmed that the pumps performed differently as a result of their different working principles. These results can be used for medical and biological applications.

In 2021, study [25] brought new innovations to the evaluation process of the equipment. They performed calibration, the uncertainty calculation, the validation of the interferometry method, and compared it to the gravimetric method. From the obtained results, the authors concluded that the gravimetric method can be used for calibration of flow devices up to 0.01 mL/h with a reasonable uncertainty value of 5% (k = 2). They reported that the interferometer method appears promising in allowing for reliable, reproducible, and more certain measurements in the microliter flow domain. In this way, future studies may make use of better measurement of syringe radius, executing including measurements of up to 0.1 $\mu$L/h, as well as testing other types of flow generators, and validating the interferometer methodology by a comparison.

In studies [35, 36], and [32] research was successful using syringe pumps. In the first one, 03 pumps were placed at the distal outlet level of the infusion line, 30 cm above and 30 cm below, to verify how variations in the height and density of the solution can influence the accuracy of the pumps. The study concludes that the position of the syringe infusion pump can influence the amount of volume infused, especially at low infusion rates. The second study worked with 07 syringe infusion pump sets, which were evaluated in an in vitro study during start-up, displacement maneuvers, and occlusion of the infusion line at pre-defined flow rates. The problems that affected all tested sets are mainly related to the working principle of the syringe of the infusion pumps and will only be partially solved through incremental improvements to the existing equipment.

The search in study [32] tested 5 syringe infusion pumps and 3 lightweight pumps. They performed a bench top comparative study with two different flow measurement methods, performed over a 2-h infusion period between amplitudes of 300 and 3000 ms. They noted that low-weight syringe infusions provide discontinuous flow with potential clinical implications for critically ill patients receiving vasoactive drugs. This study also highlights a hitherto unknown negative impact of altitude on pump function. For these reasons and problems related to syringe IBs, these variants should be considered in intravenous therapy especially for pediatric patients, to reduce medication errors triggered by changes in hydrostatic pressure and the compliance system.

In terms of equipment reliability, study [28] analyzed registration errors in typing prescriptions, using a 5-key keyboard, where results were used to determine probability distribution of the respective errors found. A percentage of typing errors of 7.13% was found for infusion volume and 16.91% for the infusion rate. These results led the authors to conclude that for the set of infusion pumps studied, by combining real empirical data with a new approach for studying usage error, this article raised several questions about medical device design, as well as lamenting the scarcity of data available to increase design priorities reducing error magnitudes and error rates.

The search of study [37] corroborates results of study [28] on the importance of having databases with information on medical equipment. The author states that, with this information, it is possible to investigate reliability and usability of the instruments reported. Because of this she carried out a study to investigate the effect of the infusion pump models on operating errors, based on stored historical data, considering pumps from four different manufacturers. To compare the equipment, she determined the failure rate and operating errors (usability measure) of each model. From the number of devices and installations, graphical analysis, by histograms, was performed to evaluate errors and usability, corrective maintenance and reliability, in addition to cost and time savings. In regards to records of infusion pumps that had been in operation for at least 1 year, and had undergone at least one corrective maintenance procedure were considered for analysis. The graph shows significant differences between manufacturers and increases over the life of the equipment. As for errors and usability, only one device showed a significant difference in relation to the others. In conclusion, the usability and reliability measures can be considered as very useful information for the acquisition of new equipment by hospitals and also for the improvement of the project by the manufacturer.

Following the area of reliability in infusion pumps, the study authors [29] verified changes in the performance of different repairable components present in the equipment. Two different types of failures were considered: those that interrupt the instrument’s operation, such as failures in indicators and relays and the occlusion of fluid flow channels and those that alert you to a fault but do not interrupt operation, such as defects in electronic circuits and audible operating signals. In this research, historical data on failures of components of infusion pumps were used, taking into account variables, such as sound signal, fairing, and battery. Graphical methods and trend tests were used for analysis. The authors review the trend tests presented in the literature and proposed a statistical test to verify the null hypothesis that the failure process follows a non-homogeneous Poisson process (PNHP). Up to this point these methods had only previously been studied with rightcensored data, and their use was proposed assuming a powerlaw density function, with the expectation–maximization (EM) algorithm applied to estimate the model parameters and compared with the modified MS. Three case studies were carried out for the mentioned components, and 80 random data sets were used for the component that beeps, 38 for the fairing, and 674 for the battery. The results showed that the parameters estimated by the EM and modified EM methods were very close.

Limitations

In preparing this systematic review, some limitations were noted. First, there was heterogeneity in the analysis of failures, characteristics of error records, models, maintenance period, and parameters evaluated which made the meta-analysis impossible. Most studies did not clearly specify the methodology, tools, and calibration of maintenance instruments, which are necessary for risk of bias and quality, assessment.

Conclusion

This review intended to investigate whether precision and reliability are aspects that have been evaluated in recent years. The research reviewed here points to the need for periodic monitoring of flow rate of the equipment, especially for low flows. This variable is essential to maintain patient safety and reduce dosage errors. In addition, it is noted that the gravimetric method is a proposal that delivers positive results in the evaluation of the flow, being an alternative for those responsible for clinical hospital engineering to utilize in routine maintenance of the devices.

Based on the findings of this review, it is noted that there is continuous need for new developments in studies that clearly adopt concepts related to reliability. In possession of the results, we verified the lack of the area applied to infusion pumps, principally studies that involve bench tests and in loco. This deficit promotes inaccurate infusions, an increase in adverse effects, and possible harm to the patient.