Intravenous infusion devices for perioperative use

INTRODUCTION — Intravenous (IV) infusion pumps are used to administer medications and fluids to many patients in perioperative and critical care settings. This topic will review the technology and use of IV infusion pumps, including simple manual flow regulators that use gravity and a roller clamp, syringe pumps, elastomeric pumps, more complex "smart" pumps that require operator programming, and closed-loop target-controlled infusion (TCI) systems. Since all IV infusion systems have potential to cause harm, the topic emphasis is on patient safety during use of these devices.

Additional aspects regarding prevention of medication errors and adverse drug events in the perioperative and other in-hospital settings are addressed in separate topics:

●(See "Prevention of perioperative medication errors".)

●(See "Prevention of adverse drug events in hospitals".)

OVERVIEW OF INFUSION DEVICES

●Types of infusion devices – The earliest (and still common) method for regulation of an intravenous (IV) infusion is a roller clamp on the IV tubing (see 'Manual flow regulators' below). This was followed by development of syringe and elastomeric pumps (see 'Syringe pumps' below and 'Elastomeric pumps' below). Early, simple pumps were set to deliver fluids at a given rate per hour and required the user to manually calculate and enter an infusion for each medication. More complex devices called "smart pumps" have been developed to improve medication safety and include a library with high-alert medications such as vasoactive agents (see 'Smart pumps' below) [1,2]. These pumps can be programmed to deliver a drug at a specified dose (eg, mcg/kg/min), have high and low flow rate alarms with a dosing error reduction system (DERS), and can log alerts to improve clinical ability to detect and prevent serious medication errors. However, the benefits of smart pumps depend on the operator's compliance with the drug library incorporated within the pump [3,4]. Different hospital locations may have different drug libraries (eg, pumps programmed for use in the medical intensive care unit [ICU] may use a different library than pumps intended for use in the cardiothoracic ICU or the operating rooms). Target-controlled infusion (TCI) systems use pharmacokinetic and pharmacodynamic mathematical modeling to maintain a target plasma concentration of a medication. (See 'Target-controlled infusion systems' below.)

●Selection of infusion devices – General considerations for selection of an appropriate infusion device include:

•For high-alert IV medications such as vasoactive agents (eg, vasopressors, inotropes, vasodilators) or insulin, we prefer smart pumps containing a standardized medication library [1].

•For delivery of small quantities of IV fluids or medications such as sedatives, opioids, or vasoactive agents at a precise rate for a short period of time, syringe pumps are an appropriate choice and are widely available for use in the perioperative setting.

•For ambulatory infusion therapies (eg, continuous peripheral nerve blockade or IV analgesic agents for postoperative analgesia), elastomeric pumps are an appropriate choice.  

•For approximate control of IV fluid infusion rates in intraoperative settings where continuous observation is possible, the simplicity of assembly and use, ready availability, and low cost of manual flow regulators make them a good choice.

RISKS FOR MEDICATION ERRORS — In the intraoperative setting, the continuous presence of an anesthesia provider means that intravenous (IV) infusions are monitored closely. Conversely, operator supervision of IV infusion pumps is more variable in most other inpatient areas of a hospital. (See "Prevention of adverse drug events in hospitals", section on 'Smart pumps' and "Prevention of perioperative medication errors", section on 'Medication infusion 'smart' pumps'.)

Programming errors — A US Food and Drug Administration review noted that dosing errors accounted for 17 percent of in-hospital medication errors [2], and the primary cause is misprogramming of IV infusion pumps [5].

●Causes – In the perioperative setting, errors during infusion of IV anesthetic agents include failure to deliver the intended agent, underdosing with possible awareness, and overdosing [6]. Although smart infusion pumps can theoretically prevent pump programming errors, they cannot prevent all error types [7]. It is still possible for a clinician to administer the wrong medication or wrong concentration, or to administer a medication to the wrong patient (see "Prevention of perioperative medication errors", section on 'Types and incidence of errors') [8]. Also, bypassing the preprogrammed dose limits in the site-specific smart pump medication library may result in inappropriate drug delivery [9], although this is sometimes necessary in the perioperative setting, particularly in emergency situations. In addition, violation of hospital policies such as using nonstandardized medication concentrations or covering labels on IV bags or tubing may contribute to drug dosing errors despite use of a smart pump. In a review of documented IV infusion errors occurring during use of a smart pump in 478 patients (1164 medications), causes included administration of medications not in the pump library (24 percent), bypassing the drug library (10 percent), incorrect infusion rate (5 percent), and failure to include IV fluid necessary to keep the vein open (ie, KVO fluid) in calculations of total infused fluids (5 percent) [10]. (See "Prevention of perioperative medication errors", section on 'Standardized concentrations of high-alert medications' and "Prevention of perioperative medication errors", section on 'Standardized labels'.)

Newer more sophisticated devices with automated features have been developed such as “hybrid” closed loop systems for insulin administration. These automatically adjust basal and correctional insulin delivery, while also allowing manual selection of prandial boluses by the patient. To avoid communication errors, it is essential that the anesthesia provider discusses the presence of such a hybrid pump and the decision to continue its use in AutoMode during the perioperative period with both the patient and, if possible, with the primary physician [11,12]. For example, hypoglycemia can occur if a subcutaneous or IV dose of insulin is administered automatically by the pump and additional dose(s) are administered manually by the provider in response to elevated blood glucose during surgery. These pumps must be continuously accessible to the anesthesia team during the procedure and not prepped into the surgical field or tucked under the drapes.

●Outcomes – Fortunately, not all medication errors result in an adverse event (figure 1) (see "Prevention of perioperative medication errors", section on 'Errors resulting in adverse medication events'). In the review of errors occurring during use of a smart pump in 478 patients cited above [10], only four of the errors were rated as having potential to cause significant harm. An observational study noted similar findings, with only 1 percent of 231 errors occurring in more than 2000 IV medication infusions that could have caused significant harm [9]. In that study, the error rate was similar with or without use of a smart pump.

●Risk mitigation – When programming a pump, checking two values (eg, micrograms per kilogram per minute and rate in milliliters per hour) may help to prevent medication errors. For example, the user might overlook an infusion rate that was set to 1 mcg/kg/minute instead of 10 mcg/kg/minute, but an infusion rate that lasts 36 hours instead of 3.6 hours would be more visible. A detailed discussion of prevention of medication errors occurring in the perioperative setting is provided in a separate topic. (See "Prevention of perioperative medication errors".)

Incompatibility of medications during multiple infusions — It is often necessary to infuse several medications through the same catheter, which may be a peripheral IV or a central venous catheter. In such cases, the compatibilities of all infused medications must be considered [13]. Physical incompatibilities may result in precipitation of the active drug, color change, gas production, or other adverse reaction. Any of these issues may significantly reduce the amount of medication administered to the patient [14]. Chemical incompatibilities can lead to drug degradation and the formation of toxic substances. The risks of infusing many common drug combinations (eg, propofol and remifentanil) through a single catheter have not been studied and should be avoided when possible [15]. Multi-lumen catheters with up to eight ports have been developed in an attempt to solve such compatibility issues [14].

Disconnection of intravenous tubing — When the limb containing an IV catheter is tucked at the patient's side or not continuously visible due to surgical draping, disconnection of the IV tubing from the IV catheter may occur, with consequent failure to administer the intended medication.

Extravasation of a continuous infusion — Unrecognized IV infiltration and extravasation of a continuous IV infusion into the tissue of a tucked limb may result in compartment syndrome [16,17]. Other factors that can exacerbate complications of extravasation include the pH and osmolarity of the infused fluids [18]. Vasopressor extravasation is also a risk factor for tissue injury. One systematic review suggests that these injuries may be more likely if the intravenous catheter is located in the antecubital or popliteal fossa [19]. A "high infusion pressure" or "obstruction" alarm indicates the need for rapid inspection of the IV site. However, extravasation typically causes only a modest change in pressure due to high compliance of the subcutaneous tissue such that no alarm is activated. Some infusion devices have been recalled due to the risk of not alarming for repeated upstream occlusion events [20]. Technology for earlier detection of extravasation is currently under development [21].

Cybersecurity risks — Although the ability to connect IV infusion pumps to the electronic health record (EHR) increases available data, provides a decision support resource, and may lead to improved patient outcomes, this interface incurs risks in the absence of stringent cybersecurity measures. Hackers can theoretically steal control of a device to disable hospital networks or cause harm to many patients as part of a bioterrorist attack [22,23]. Ransomware hackers have previously attacked entire hospital systems (eg, the WannaCry ransomware attack in the United Kingdom's National Health Service [NHS]) [24].

Important cybersecurity precautions include access management and cloud storage. However, there are scant data to inform development of comprehensive guidelines and standardized best practice measures [25]. Although a detailed review of protective measures is beyond the scope of this review, information technology managers should ensure that new systems and networks are correctly installed and configured. Healthcare organizations should continuously monitor computer, device, and application use to detect suspicious activities. These steps will increase the likelihood that security problems are identified and addressed before they can harm a patient [26].

MANUAL FLOW REGULATORS — The oldest way to regulate intravenous (IV) infusions is to manually provide simple, adjustable resistance with a roller clamp on the IV tubing. Another example is use of an incorporated Dial-a-Flow apparatus, which provides a mechanism to change the resistance to flow from partial to complete occlusion by compressing a longer length of IV tubing inside the device. Typically, numbers on a Dial-a-Flow apparatus range from zero to 100, with zero representing complete tubing occlusion, while 100 is the lowest resistance setting allowing for maximum flow. Notably, Dial-a-Flow is an analog device that can be set to an approximate flow rate and changed incrementally.

Macrodrop and microdrop infusion sets are available, including sets with 10, 15, 20, or 60 drops constituting 1 mL. The approximate flow rate can be determined by counting the drops in the infusion set's drip chamber.

●Adjusting the flow rate

•General considerations – The flow rate of an infusion with a manual flow regulator is adjusted by verifying that the pressure gradient is adequate and then adjusting the resistance in the tubing with a roller clamp (figure 2). The higher the gradient the higher the flow. However, as the fluid in the IV bag flows into the patient, the pressure gradient may decrease. Therefore, as height of the fluid in the bag decreases, the rate of fluid flow into the patient also decreases. Without continuously adjusting the roller clamp for a lower resistance, the flow of fluid will gradually decrease and eventually stop when the pressure gradient reaches zero.

Other components affecting flow rate can be realized by examining Poiseuille's Law (figure 2):

Q=∆P∙π∙r48∙h∙L or Q = ΔP/R

Note that:

-Q = Flow

-ΔP = the pressure gradient at the bottom of the fluid column

-r is the radius of the tubing (note that radius is raised to the fourth power). When the radius becomes larger, resistance decreases and flow increases.

-h is the viscosity which increases resistance to flow and decreases flow. For example, blood has a higher viscosity than crystalloid fluid such as normal saline. Regardless of the type of infusion solution (or red cell transfusion), the relationship between pressure and flow rate in the entire infusion system is nearly linear [27].

-L is the length of the tubing. The longer the tubing the lower the flow.  

•Specific techniques to increase infusion rate

-The IV pole can be raised or the patient bed can be lowered to increase the gradient.

-The fluid bag can be placed into a pressure bag, although this carries the risk of air embolism unless all air in the infusion system is carefully eliminated prior to administration. Furthermore, pressurizing the infusion solution bag increases risk of extravasation if the IV is in a tucked limb. (See 'Extravasation of a continuous infusion' above.)

-A larger bore IV catheter can be selected since the catheter itself produces more resistance, and a slow rate may be imposed by a small-bore catheter. Infusions delivered at a pressure >200 mm Hg benefitted from increasing IV catheter size from 18 to 14 gauge [28].

-A low-resistance drip chamber can be selected. At pressurized infusion with a large-bore IV catheter, the standard drip chamber becomes the limiting component and imposes the largest resistance to flow [28].

-An additional IV catheter can be inserted as an effective way to increase the infusion rate when gravity is used to drive the infusion [28].

●Advantages and disadvantages of manual flow regulators

•Advantages

-They are simple to assemble and easily adjusted to change the IV tubing resistance to allow increases or decreases in the rate of IV infusion.

-There is no need for batteries, electrical power, or manual programming.

-Low cost, intuitive use, and ready availability makes it easy to control approximate fluid infusion rates in settings where continuous observation is possible (eg, the intraoperative setting).

•Disadvantages

-Potential inaccuracies including the potential for infusion of the entire bag of IV fluid (eg, 1000 mL) if the provider is distracted. Gravity-fed administration systems are generally inaccurate because a variety of factors may affect the flow rate:

Changing the height of the bag relative to the patient's heart will change the pressure differential and thereby the flow rate. For example, lowering the operating room table will increase the flow rate.

As fluid is infused into the patient from the bag, the height of the fluid in the bag decreases, which also decreases the pressure gradient and the flow rate (figure 2).

The resistance of the drip chamber is important for achieving the range of the infusion set. The diameter of the IV catheter is the rate-limiting component of the infusion set. Poiseuille’s Law states that resistance to flow varies with the fourth power of the diameter, so even a small decrease in bore size can significantly increase resistance and allow less flow than a large-bore catheter [28]. One study evaluated drop rates in a variety of infusion sets (macrodrop and microdrop) [29]. Less than 15 percent of observations were within 10 percent of the desired flow rate after initially setting the roller clamp, while only 21 percent were within 20 percent of the desired flow rate. Overall, there were substantial differences between the observed and desired flow rate, indicating that manual regulation is less accurate than infusion pumps.

A disadvantage for the Dial-a-Flow apparatus is its inability to deliver a calibrated flow rate. The user must understand the limitations of the device as there is a high risk for overdosing or underdosing, potentially causing patient harm. For this reason, manual flow regulators are unsuitable for administration of potent medications such as vasoactive agents (eg, vasopressors, inotropes, vasodilators).

•Documentation issues

-Automatic documentation of changes in rate of administration of a medication or fluid (eg, in an electronic health record [EHR]) is not possible since flow rates are not monitored. Potential solutions for this problem include devices that monitor the drop rate through the infusion chamber (eg, DripAssist). Such battery-powered devices display the flow in mL/minute, as well as the total volume infused. They may be useful in certain situations (eg, disaster scenarios with a power outage).

-Infusions administered via any manual flow regulator cannot be left unattended. However, such infusions can be used to control approximate fluid infusion rates in intraoperative settings with continuous observation.

SYRINGE PUMPS — Syringe pumps allow a healthcare professional to administer small amounts of fluid at a precise rate that is set by the user. The accuracy of the pump depends on selection of the correct syringe during pump programming, but most pumps can automatically identify the size of syringe if the user correctly enters the name of the syringe manufacturer.

●Advantages and disadvantages of syringe pumps [30]:

•Advantages

-Syringe pumps are widely available for use in the perioperative setting to deliver small quantities of fluids or medications (eg, sedatives, general anesthetics, vasoactive agents) at a precise rate.

-Most commercially available devices include integrated safety features such as a drug library. In one study of 133,601 in-hospital medication infusions, the institutional medication library was used to set dose rates for 92.8 percent of syringe pump infusions [31]. The most frequently administered medication classes were vasoactive agents, followed by sedatives [31].

•Disadvantages

-Delays in starting an infusion

If there is a gap between the syringe driver and the plunger, delivery of the medication to the patient may be delayed. This brief delay may not be clinically significant when an adult is receiving an antibiotic but may be critical when beginning an urgently needed fluid infusion in a neonate or a vasoactive drug in a hemodynamically unstable patient. The time from beginning the infusion to reaching a steady-state infusion rate depends in part on the set flow rate (0.1, 0.5, or 1 mL/hour), and in part on the syringe volume size (eg, 10, 20, 30, or 50 mL). Delays as long as 20 to 75 minutes have been noted if the syringe size is large and the set flow rate is low [32,33]. The shortest delays (approximately four minutes) occurred with smaller syringes (eg, 10 mL) and flow rates of at least 1 mL/hour. Clinicians have used "workarounds" for this problem such as doubling the flow rate for a few minutes before connecting the infusion to the patient, but this practice is not recommended as it may result in an overdose if the clinician is distracted and fails to reset the pump to the correct rate.

A delay may also occur due a hysteresis effect (ie, lag time) as the pump's motor must turn multiple gears during startup.

-Parameter entry errors – For example, the drug library may contain a drug that should be infused as mcg/kg per hour, but the user can manually change the pump to infuse mL/hour or mcg/minute instead. Furthermore, it may not be clear which units of measurement the user is expected to enter (weight in pounds may be entered when the infusion pump requires weight in kilograms).

-Alarm malfunction – The infusion pump may fail to generate an audible alarm for a critical problem, such as an occlusion (eg, clamped tubing or an infiltrated IV line) or the presence of air in the infusion tubing. Conversely, the pump may generate an occlusion alarm in the absence of an occlusion. Also, an alarm indicating low battery charge may not be displayed in time for a user to prevent pump shut-off during a critical infusion while a patient is in transport [30].

-Unclear warnings – Warning messages may be unclear. An example is the message "Volume in the syringe is inadequate to deliver the programmed dose – PRESS CONFIRM;" it is unclear if the user is confirming the desired infusion settings or his or her understanding of the warning message itself.

-Temporary flow increases may occur when the height of the pump is increased, or flow decreases when the pump is displaced downward [34].

ELASTOMERIC PUMPS — Elastomeric pumps have an elastic chamber or balloon that stretches to store energy and pressure (typically 260 to 520 mmHg) as the chamber is filled with the fluid or medication to be administered. The balloon then returns to its original form as it pushes the liquid out through the tubing. The flow controller is used to restrict flow as desired, similar to setting the resistance on a manual flow regulator. (See 'Manual flow regulators' above.)

Elastomeric pumps rely upon the physical characteristics of the balloon to drive the infusion and do not use batteries or any other source of electric power. The figure illustrates the assembly of an elastomeric pump, depicting an initially deflated elastomeric balloon reservoir covered by a hard outside shell that protects the balloon, the filling port, the delivery tube that includes a filter, and the flow controller (figure 3).

Elastomeric pumps were designed to provide ambulatory infusion therapies, such as chemotherapy and postoperative analgesia [35]. They are commonly used for continuous peripheral nerve blockade or IV analgesic agents (eg, methadone, tramadol, dexketoprofen, ondansetron) [36-38].

Advantages and disadvantages of elastomeric pumps include:

●Advantages

•Elastomeric pumps are simple to use and have been associated with fewer human errors during setup and fewer technical difficulties during use compared with other pumps [35,39].

•There is no need for batteries or electrical power.

•For ambulatory patients, specific advantages include:

-Portability for use to deliver postoperative analgesia at home in the immediate postoperative period [38].

-Single-use disposable pumps that can be refilled several times before being discarded, with maintenance of delivery rate and performance after repeated filling [40].

-Infusion duration that may last for several hours up to seven days.

-Overall safety. In one study, fewer than 2 percent of patients receiving home-based analgesic administration had device-related adverse effects or catheter-related complications (eg, phlebitis, extravasation, device dysfunction) [38].

●Disadvantages

•Accuracy and consistency

-Accuracy and consistency of delivery rate are generally poor. Accuracy is ±15 percent. For example, a pump with a nominal 5 mL/hour actually delivers in the range of 4.25 mL/hour to 5.75 mL/hour [40].

-Variations in actual versus set flow rate may occur due to many factors [37]:

-Initial infusion rates are generally faster than the specified flow rate, as illustrated in the figure (figure 4) [40].

-Actual pressure exerted by the elastomeric balloon is determined by the filling volume.

-Use of catheters greater than 22 gauge are necessary to maximize accuracy.

-The vertical height of the pump in relationship to infusion site may send positive or negative back pressure depending on whether the pump is above the IV catheter or wound site (positive pressure) or below this site (negative pressure).

-Flow rate is inversely proportional to the viscosity of the administered fluid. For calibrations of flow, 5% dextrose in water (D₅W) is used.

-Temperature affects elastomeric pump function, with variations from 2 to 3 percent for each 1°C. Devices are most accurate at 92°F or 33.3°C.

-Partial filling of an elastomeric pump can alter the exerted internal pressure and consequently the flow rate [37].

•Difficulties with monitoring and alarms

-The amount of drug or total volume delivered cannot be measured, a particular disadvantage in pediatric patients [41].

-There are no alarms on elastomeric devices.

•Special considerations with antibiotics – Degradation of antibiotics within the elastomeric chamber may occur [39,42,43]. Nevertheless, with changes of the elastomeric pump every 24 hours, continuous antibiotic therapy has been administered for 13 days to manage persistent osteoarticular infections, with a reported success rate of 96 percent [40].

SMART PUMPS — So-called "smart," pumps for IV medication infusion ensure that the programmed infusion rate for a given medication is within pre-existing limits using an institutional standardized medication library approved by the institution's pharmacy department. Smart pump technology can be incorporated into any electronic pump, including syringe pumps or peristaltic pumps. Audiovisual feedback is provided if the user attempts to program infusion parameters that are outside the predetermined dosing limits or use of incorrect concentrations or duration thresholds. This dose error reduction system (DERS) is programmed with "hard" limits determined by the maximum and minimum safe doses that cannot be bypassed by the user programming the pump.

Smart pumps also have "soft" limits usually determined by the most commonly used infusion rates for a medication. Exceeding a soft limit causes a warning to be displayed noting that the dose is too high or too low, but the user is permitted to proceed with the desired programming and start the infusion after acknowledging the warning. As an example, the figure shows a DERS system, with a mean flow of 10 mg/hour (figure 5) [44]. The normal therapeutic range for the flow is 5 to 30 mg/hour. If 30 mg/hour is not sufficient, the clinician may bypass the soft upper limit and increase the flow to 40 mg/hour. If, however, the flow rate is set to greater than 60 mg/hour, the pump will reject the setting because the hard upper limit has been exceeded. Opportunities for improvement include approaches to minimize "workarounds" that bypass safety limits, but without increasing the number of unnecessary (false) and distracting alarms [7,9]. A study of patient-controlled analgesia (PCA) pumps concluded that device-related errors are relatively uncommon, occurring in less than 0.2 percent of patients, and that the most common error is underflow, resulting in inadequate analgesia. The authors found no instances of patient mortality [45].

In practice, the safety and efficacy of smart pumps depends primarily on the operator's use of the medication library incorporated within the pump, and acceptance of the suggested dosing limits for most patients (rather than manual programming that allows bypass of these safeguards) (figure 6) [3,4,7]. Most smart pumps can be programmed to revert to a "basic" mode (eg, mL/hour). Clinicians may be tempted to use this mode because it is familiar, but doing so bypasses the safety features incorporated in the drug library.

In the operating room, smart pumps are frequently used to administer intravenous (IV) anesthetic agents and vasoactive medications (eg, vasopressors, inotropes, vasodilators). These agents must be continuously titrated to achieve the desired effect. Ideal features for a smart infusion pump in the perioperative setting include operator ability to rapidly change the flow rate with minimal keystrokes. For example, nine or more characters may be required for entry of the patient's identification number and other information such as weight, followed by multiple keystrokes to enter the medication name, concentration, and desired initial dose. Error rates increase with the number of required keystrokes [7]. In some perioperative circumstances, adjustment of the flow rate to provide doses that are outside the device's preprogrammed limits may be necessary. However, in all such instances, appropriate warnings should be displayed on the device.

Advantages and disadvantages of smart pumps include [7]:

●Advantages

•The potential to detect and prevent serious medication errors and improve patient safety. In a 2014 systematic review, use of smart pumps reduced programming errors but did not eliminate them [7]. This study found that compliance with drug libraries and smart pump limits was the best way to prevent drug errors. Drug library hard limits were the most effective way to minimize medication errors, while soft limits were not as effective because users override them.

•Inclusion of features that can usually identify the source of medication errors due to retention of a log of all alerts. In one study, smart pump alert data were collected to determine which medications were associated with the most and least clinically meaningful pump alerts [46]. These data were then used to optimize the drug library limits and also to decrease clinician "alert fatigue."

•Other technology to detect problems such as occlusion, infiltration, siphoning, and air bubbles in the IV fluid is incorporated within such devices [47].

•Inclusion of a barcode-assisted medication administration scanning system is possible [48]. Scanning of barcodes may include those on medications, the patient's wristband identification, and the clinician's identification badge.

●Disadvantages

•Possible user errors

-Pump setting errors that may go undetected (eg, wrong rate, wrong dose, wrong concentration, wrong medication)

-Use of the wrong medication library

-Overriding soft limits

-Compliance issues (ie, workarounds)

•Limitations for individual patients

-Limits are set for typical patients. It is not possible to predict the needs of every patient and clinical circumstance. For this reason, many IV smart pumps allow the users to bypass the DERS low end limits and use manual programming as needed. However, this ability to bypass the pump may also be a disadvantage that can cause medication errors in some instances.

-Smart pumps are not assigned to individual patients and are not associated with intended therapy for an individual patient. Thus, they cannot intercept wrong patient medications or prevent wrong drug selection.

•Medication library issues

-The medication library cannot contain every medication that may be used in the intraoperative setting, or may contain only very narrow parameters for a given medication, so that manual programming may be necessary. In such cases, the user must take extra care to ensure that the infusion rate is correct.

-Implementation of the use of smart pumps throughout an institution requires consolidation of the hospital medication formulary and elimination of use of unapproved abbreviations. Typically, there is a necessary "learning curve" for clinicians during initial clinical use. Nevertheless, follow-up at one year after implementation noted a reduction in pump-related dosing errors from 41 to 22 percent in one study, with more minor reductions in wrong dose and incorrect infusion rate errors [49].

TARGET-CONTROLLED INFUSION SYSTEMS — Target-controlled infusion (TCI) systems are computer-assisted IV infusion pumps that use pharmacokinetic and pharmacodynamic mathematical modeling to maintain a user-designated target concentration at an effect site (typically the brain) [50,51]. When using a TCI system, the clinician enters a desired target concentration for an anesthetic or other agent and information such as the patient’s height, weight, and age. The computer then calculates the amount of the agent required to achieve the target concentration at the effect site, and directs an infusion pump to deliver the calculated boluses or infusions. Subsequently, the computer constantly recalculates how much drug is in the tissues and how that influences the amount of drug required to achieve the desired target concentration at the effect site. Calculations include the drug's pharmacokinetics as well as patient covariate data (eg, age, weight).

Mathematical modeling for TCI infusions must account for redistribution and elimination of the medication into multiple compartments. A three-compartment model is typically used (figure 7) [52]. The central compartment (V1) represents the plasma concentration as the pump injects the medication intravenously, and one peripheral compartment is largely comprised of fat tissue (V2), while the other peripheral compartment represents lean tissue concentration (V3). Calculations of rate and direction of drug movement among these compartments occur continuously throughout the period that the medication is administered in order to maintain a targeted effect site concentration.

In general, a TCI pump will initially deliver a bolus dose to rapidly achieve the desired end-organ concentration, followed by administration of an infusion or smaller bolus doses to maintain the target concentration. Hysteresis (ie, delay in the monitored estimate of drug effect) typically occurs [53]. In one study of propofol titration using TCI technology, improved estimates of propofol target effect site concentration were achieved by adding a lag time of approximately 50 seconds to the pharmacodynamic model for two monitors of drug effect (bispectral index [BIS] electroencephalography [EEG] and qCON index) [53]. A TCI system typically reduces the infusion rate gradually over time, based on its calculations. However, the desired propofol target effect site concentrations titrated using BIS values are variable [54]. For example female patients and patients with body-mass index (BMI) >26 kg/m² typically have lower BIS values [55]. Furthermore, differences in EEG characteristics caused by different anesthetic drugs results in variability.

These concepts are intuitive for anesthesiologists since they are similar to techniques used for delivery of inhalation anesthetics to achieve a desired end-tidal concentration, with a corresponding brain concentration of the anesthetic after equilibration has occurred. (See "Inhalation anesthetic agents: Clinical effects and uses", section on 'Induction of general anesthesia' and "Inhalation anesthetic agents: Clinical effects and uses", section on 'Maintenance of general anesthesia (all inhalation agents)'.)

In the perioperative setting, TCI techniques have been used to induce and maintain general anesthesia or to provide computer-assisted personalized sedation [51,56,57]. However, there are potential inaccuracies for commonly used pharmacokinetic/pharmacodynamic models for plasma concentrations of anesthetic agents [58]. Notably, TCI devices have not been approved in the United States [56,59,60].

Advantages and disadvantages of TCI systems include:

●Advantages

•Theoretically, TCI allows predictable induction and maintenance of general anesthesia or sedation, as well as predictable recovery from anesthetic effects when used in appropriate patients and settings [6,61-68].

•Improvements in performance and safety of TCI systems are likely such as incorporation of closed-loop controls based on monitoring of pharmacodynamic end-points such as processed electroencephalographic (EEG) parameters or changes in the hemodynamic profile (eg, mean arterial pressure and heart rate) [6,62,63,66,69-72]. Such pharmacodynamic monitoring could supplement TCI pharmacokinetic modelling of effect-site concentrations. Closed loop systems are being developed that require continuous measurement of a biosensor output (eg, EEG), with continuous adjustments of the infusion rate to maintain a target value.

●Disadvantages

•Intersubject variability in the pharmacokinetic and pharmacodynamic parameters used in models to calculate the drug concentration at the effect site has limited the utility of TCI systems [6,57,61,73,74]. In particular, the following patient covariates affect modeling:

-Obesity is commonly associated with higher than anticipated plasma concentrations of propofol and other agents with use of standard models for TCI pumps [73,75,76]. Models that incorporate obesity are accurate only in patients with obesity, but risk underdosing for other patients [75].

-Underweight individuals may experience either underdosing or overdosing with available models [77].

-Children may have different pharmacokinetic and pharmacodynamic profiles than adults. Use of TCI for propofol anesthetic administration can lead to higher propofol dosing in pediatric patients, with or without prolonged recovery time [52,78,79].

-Most TCI models do not compensate for patient comorbidities [6,80].

-Most TCI models do not compensate for drug-drug interactions or synergy when multiple medications are administered [6].

•The need to develop multiple pharmacokinetic models for each anesthetic agent in previous research studies of TCI system use has limited the clinical utility of these systems [57,62,81]. In the future, selection of the appropriate model for different patient groups (eg, patients with obesity, pediatric patients) will likely be automatic during device programming for an individual patient [57,59].

LIKELY FUTURE DEVELOPMENTS — Continuing developments for intraoperative IV infusion technology to improve safety and ease of use include clinical integration with the anesthesia information management system, thereby incorporating use of clinical monitoring devices (pulse oximeter and capnography). For example, in an awake patient, if oxygen saturation decreases or carbon dioxide increases beyond certain preset limits during infusion of a medication such as an opioid or sedative agent, the infusion pump(s) would automatically stop, and alarms in the system would be activated [82].

The next generation of infusion management systems may be able to directly interface with the patient's electronic health record (EHR) and computerized physician order entry (CPOE) system, ideally with barcode-assisted medication administration. These wireless bidirectional connections will allow decision support for preprogramming of IV infusion devices to minimize risk of errors, including calculation of appropriate infusion rates based on factors such as patient weight and known standardized drug concentrations in an institution's premixed infusions [50,83,84]. In addition, electronic transmission of infusion information obtained from the pump will occur automatically to document titrations of infusion rates, thereby allowing continuous recording of medication dosing within the EHR [5,50]. The resulting electronic medication administration records would be used to target errors that occur in drug-dispensing transcription and administration within an institution. (See "Prevention of adverse drug events in hospitals", section on 'Electronic medication administration record' and "Prevention of adverse drug events in hospitals" and "Prevention of perioperative medication errors", section on 'Clinical decision support'.)

Other likely future developments include:

●Elimination of most false ("nuisance") alarms that may decrease users' sensitivity to all alarms.

●Improvements in pharmacokinetic modeling and supplemental pharmacodynamic monitoring for target-controlled infusion (TCI) systems. (See 'Target-controlled infusion systems' above.)

●Systems that take advantage of alternate routes for medication administration (rather than IV delivery). Examples include intranasal, pulmonary, buccal mucosal, and intra-articular controlled-release drug therapy, as well as transdermal medication delivery systems [85].

●Advanced infusion technologies such as a new “cylinder” infusion that can control infusion rates as accurately as a syringe pump, but allows the use of larger containers of fluid and/or drugs [34].

SUMMARY AND RECOMMENDATIONS

●Risks for medication errors – During perioperative use of intravenous (IV) infusion devices include:

•(See 'Programming errors' above.)

•(See 'Incompatibility of medications during multiple infusions' above.)

•(See 'Extravasation of a continuous infusion' above.)

•(See 'Cybersecurity risks' above.)

●Manual flow regulators – Roller clamp providing adjustable resistance on the IV tubing (or an incorporated Dial-a-Flow apparatus) (figure 2). (See 'Manual flow regulators' above.)

•Advantages – Provides simplicity and easy adjustability to allow changes in IV infusion rate; absence of need for batteries, electrical power, or manual programming; low cost; and ready availability.

•Disadvantages – Multiple factors affecting flow rate (eg, changing the height of the bag relative to the patient's heart or the height of the fluid in the bag as fluid is infused, diameter of the IV catheter), and inability to deliver a calibrated flow rate, leave the IV infusion unattended, or accurately document changes in rate of fluid administration.

●Syringe pumps – Small syringe infusion pumps. (See 'Syringe pumps' above.)

•Advantages – Ability to administer small amounts of medication or fluid at a precise rate set by the user. Other advantages include integrated safety features such as a drug library.

•Disadvantages – Delays in starting an infusion due to any gap between the syringe driver and the plunger and a hysteresis effect as the pump's motor turns to start multiple gears during startup.

●Elastomeric pumps – Pumps with a balloon that stretches to store energy and pressure as the chamber is filled with fluid, then returns to its original form as it pushes liquid out through the tubing (figure 3). (See 'Elastomeric pumps' above.)

•Advantages – Ease of use with fewer human errors during setup and fewer technical difficulties compared with other pumps, as well as absence of need for batteries or electrical power, and portability for use at home (eg, to deliver postoperative analgesia).

•Disadvantages – Poor accuracy and inconsistent flow rate since many factors may cause variations in actual versus set flow rate.

●Smart pumps – "Smart," pumps ensuring that the programmed infusion rate for a given medication is within the preset limits of a standardized medication library (figure 5). (See 'Smart pumps' above.)

•Advantages – Potential to detect and prevent serious medication errors, as well as incorporated technology to detect problems such as occlusion, infiltration, siphoning, and air bubbles in the IV fluid.

•Disadvantages – Dependence on the operator's compliance with institutional medication library limits and correct use of the pump in an individual patient (figure 6).

●Target controlled infusion (TCI) systems – Computer-assisted IV infusion pumps that use pharmacokinetic and pharmacodynamic mathematical modeling to maintain a user-designated target concentration at an effect site (typically the brain) (figure 7). (See 'Target-controlled infusion systems' above.)

•Advantages – Predictable induction and maintenance of general anesthesia or sedation, as well as predictable recovery from anesthetic effects.

•Disadvantages – Intersubject variations (eg, overweight or underweight status, comorbidities, age) that affect pharmacokinetic and pharmacodynamic parameters used in models to calculate drug concentration at the effect site, and the need to compensate for drug-drug interactions or synergy when multiple medications are administered.

●Future developments – Continuing developments to improve safety and ease of use include integration with the anesthesia information management system and hospital electronic health record, as well as improvements in pharmacokinetic modeling and supplemental pharmacodynamic monitoring for TCI systems, alternate routes for medication administration (rather than IV delivery), and elimination of most false ("nuisance") alarms. (See 'Likely future developments' above.)