Article Text

Local anesthetic dosing and toxicity of pediatric truncal catheters: a narrative review of published practice
  1. Michael Fettiplace1,2,
  2. Lana Joudeh1,2,
  3. Brittani Bungart1,2 and
  4. Karen Boretsky2,3
  1. 1 Department of Anesthesia, Critical Care and Pain Medicine, Massachusetts General Hospital, Boston, Massachusetts, USA
  2. 2 Department of Anaesthesia, Harvard Medical School, Boston, Massachusetts, USA
  3. 3 Department of Anesthesiology, Critical Care and Pain Medicine, Boston Childrens Hospital, Boston, Massachusetts, USA
  1. Correspondence to Dr Michael Fettiplace, Department of Anesthesia, Critical Care and Pain Medicine, Massachusetts General Hospital, Boston, Massachusetts, 02114, USA; fettiplace{at}


Background/importance Despite over 30 years of use by pediatric anesthesiologists, standardized dosing rates, dosing characteristics, and cases of toxicity of truncal nerve catheters are poorly described.

Objective We reviewed the literature to characterize dosing and toxicity of paravertebral and transversus abdominis plane catheters in children (less than 18 years).

Evidence review We searched for reports of ropivacaine or bupivacaine infusions in the paravertebral and transversus abdominis space intended for 24 hours or more of use in pediatric patients. We evaluated bolus dosing, infusion dosing, and cumulative 24-hour dosing in patients over and under 6 months. We also identified cases of local anesthetic systemic toxicity and toxic blood levels.

Findings Following screening, we extracted data from 46 papers with 945 patients.

Bolus dosing was 2.5 mg/kg (median, range 0.6–5.0; n=466) and 1.25 mg/kg (median, range 0.5–2.5; n=294) for ropivacaine and bupivacaine, respectively. Infusion dosing was 0.5 mg/kg/hour (median, range 0.2–0.68; n=521) and 0.33 mg/kg/hour (median, range 0.1–1.0; n=423) for ropivacaine and bupivacaine, respectively, consistent with a dose equivalence of 1.5:1.0. A single case of toxicity was reported, and pharmacokinetic studies reported at least five cases with serum levels above the toxic threshold.

Conclusions Bolus doses of bupivacaine and ropivacaine frequently comport with expert recommendations. Infusions in patients under 6 months used doses associated with toxicity and toxicity occurred at a rate consistent with single-shot blocks. Pediatric patients would benefit from specific recommendations about ropivacaine and bupivacaine dosing, including age-based dosing, breakthrough dosing, and intermittent bolus dosing.

  • Drug-Related Side Effects and Adverse Reactions
  • Nerve Block
  • Pain, Postoperative
  • Postoperative Complications

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, an indication of whether changes were made, and the use is non-commercial. See:

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Pediatric anesthesiologists use truncal peripheral nerve catheters routinely for perioperative pain control for a variety of surgeries. Continuous paravertebral nerve blocks were introduced for adult use in 1979,1 with the practice extended to infants and children by 1992.2–6 Transversus abdominis plane (TAP) catheters followed in 2010. Pediatric dosing recommendations for bupivacaine infusions arose in 1992, following reports of local anesthetic systemic toxicity (LAST) from continuous neuraxial infusions. These recommendations advised 0.2–0.5 mg/kg/hour in infants less than 6 months to 0.4–0.75 in children equal to or older than 6 months.7–9 Formal dosing recommendations for peripheral nerve catheters published in 2018 by a workgroup representing the American Society of Regional Anesthesiology (ASRA) and the European Society of Regional Anesthesiology (ESRA)10 advised the following:

Continuous infusion of local anesthetic for peripheral nerve and fascial plane blocks can be safely and successfully performed with 0.2% ropivacaine or bupivacaine using an infusion dose of 0.1 to 0.3 mg/kg per hour. (Evidence B3)

However, evidence-based dosing recommendations remain uncertain,11 with package inserts,12 13 textbooks,14–16 experimental data,17 18 and expert opinions7 10 11 providing wildly different numbers. With the addition of new sites for truncal peripheral nerve catheters, we chose to review the literature to characterize the dose, administration details, and complications of two early adopted continuous truncal blocks: the paravertebral block (PVB) and TAP block. Data from the Pediatric Regional Anesthesia Network (PRAN) exist to describe some of these patterns19–21 but primarily reflect practices within the USA. Herein we describe dosing patterns and toxicity in patients less than 18 years old in the published literature.


For full methods, please see the online supplemental material. A literature search was performed on PubMed, EuropePMC, Embase, Web of Science, and Google Scholar to identify publications between 1987 and 2021 (based on initial search in 2021) that included catheters or infusions in the PVB or TAP space dosed with bupivacaine or ropivacaine. Any study design (including technical reports, case reports, case studies, observational studies, retrospective cohort studies, and randomized trials) with dosing details intended for ≥24 hours of use were included. Exclusion criteria were alternative local anesthetics for infusion (eg, lidocaine, mepivacaine, and chloroprocaine), alternative infusion sites (eg, erector spinae), unavailable dosing, review papers, and animal studies. Non-English language papers were translated using Google Translate. No primary author was contacted for additional data. Papers and extraction of data (online supplemental table S1) were managed in Covidence (Melbourne, Australia). Median dosing and weight values with ranges were prioritized over mean values with SD. The data were split into adult catheters (presented in the accompanying paper: Bungart et al.22) and pediatric data in the current paper.

Supplemental material

Our goal was to assess and quantify catheter dosing including bolus, infusion, and 24-hour cumulative dose. See ‘Data synthesis’ in online supplemental methods for details of data processing and aggregation. We also analyzed data based on age (less than 6 months or greater than 6 months) and drug (ropivacaine or bupivacaine). The 6-month cut-off is based on pharmacokinetic data demonstrating increased local anesthetic accumulation in infants less than 6 months.7 21 In cases of ambiguity in dosing based on age, we assessed based off the median or mean age. We also assessed for toxic events and toxic blood levels associated with catheter infusions of greater than 24 hours.


Extraction and demographics

We screened a total 1344 abstracts (online supplemental figure S1) with an initial screen of PubMed and EuropePMC and follow-up screen of Embase, Web of Science, and Google Scholar. We identified a total of 46 publications,3 5 23–66 representing data from 945 patients (online supplemental tables S2–S4). An additional 220 papers contained adult data and is discussed in the accompanying paper (Bungart et al.22)Age-based weights were added for two patients (16.766 and 15 years old).67 A patient from Bollag et al 36 received two ‘as-needed’ doses in 24 hours and was included in the 24-hour dose but excluded from infusion calculations. Of the 46 publications, some included multiple infusions, which we treated separately based on weight and/or dosing with 63 infusions included for analysis. For data processing, Larsson et al 60 and Elder et al 29 were excluded as the number of patients under age 18 was unclear. Full data on bolus and infusion dosing are available in online supplemental tables S5–S8. Paper-specific details and patient-specific details are presented in online supplemental tables S3 and S4, respectively.

Initial use of PVB catheters was limited to UK and Europe between 1992 and 1997 with a break in reports until 2010, where the practice was picked up worldwide following the advent of ultrasound guided blocks in children.32 68 The majority of the data arose from the USA (55% of cases) but included a minority of patients under 6 months (7 out of 144, 5%). Data were primarily observational (52%) or retrospective cohort (33%) in nature. A paucity of patients came from randomized controlled trials (RCTs) (7%), but of those, 93% came from two studies out of Egypt37 and Iran,31 respectively. As anticipated, thoracic cases predominated (62%) primarily paravertebral (95%), unilateral (61%), continuous delivery (99%), and weight-based infusion dosing (82%)

Initial bolus dosing

Bolus doses used available concentrations (eg, 0.5% or 0.25% for bupivacaine and 0.5% or 0.2% for ropivacaine) with primarily weight-based dosing (online supplemental tables S2, S4). Median bupivacaine bolus doses were all below 2.5 mg/kg (table 1 and figure 1A), whereas ropivacaine doses extended up to 3 mg/kg with a few doses of up to 5 mg/kg (figure 1B). Bupivacaine bolus doses were similar above and below 6 months, whereas ropivacaine bolus doses were limited to 2.2 mg/kg in the ropivacaine group under 6 months.

Table 1

Bolus dosing characteristics

Figure 1

Median dosing of catheters. (A) Bupivacaine bolus dosing (n=294), (B) ropivacaine bolus dosing (n=466), (C) bupivacaine infusion dosing per hour (n=423) and (D) ropivacaine infusion dosing per hour (n=521).

The most frequent bupivacaine bolus was ~1.5 mg/kg and included combinations of 0.1% bupivacaine and median dose of 1.37 mL/kg (n=83),54 0.25% bupivacaine at ~0.5 mL/kg (n=48),5 24 25 32 33 37 56 59 and 0.5% at 0.3 mL/kg.42 The most common ropivacaine bolus was ~2.5 mg/kg and included 0.5 at 0.5 mL/kg.40 48 Bupivacaine boluses were recorded up to a maximum of 3.13 mg/kg,54 and ropivacaine bolus doses were recorded up to 5 mg/kg with bilateral PVB catheters, dosed with 0.5 mL/kg of 0.5% ropivacaine per side (age down to 7 years).40 48 Bolus doses of ropivacaine in patients over 6 months were higher in the USA (median 2.5 mg/kg, range 0.8–5.0 mg/kg; n=367) (online supplemental table S9) than in the rest of the world (median 1.5 mg/kg, range 0.6–1.5 mg/kg; n=82).

Infusion dosing (combined continuous and intermittent bolus)

Infusion doses also used available concentrations with 1:1 dilutions (eg, 0.5%, 0.25%, 0.125% and 0.625% for bupivacaine and 0.5% or 0.2% for ropivacaine) with primarily weight-based dosing (table 2). Despite larger variability, infusion doses were lower in bupivacaine (figure 1C) with a median of 0.25 mg/kg/hour under 6 months and 0.33 mg/kg/hour over 6 months; median infusion doses of ropivacaine were 0.5 mg/kg/hour for both over and under 6 months (table 2 and figure 1D). For bupivacaine, there was an equal spread of doses between 0.2 mg/kg/hour and 0.4 mg/kg/hour with occasional doses of 0.5 mg/kg/hour or above. The maximum bupivacaine infusion of 1.0 mg/kg/hour was published in 19923 with no subsequent publications with rates this high. In contrast, the most common ropivacaine infusion was 0.5 mg/kg/hour using ropivacaine 0.2% at 0.25 mL/kg/hour39 40 44 48 52 57 62 and 0.5% at 0.1 mL/kg/hour (n=34).31 International bupivacaine dosing was more heterogeneous (range 0.08–1.0 mg/kg/hour) than dosing in the USA (range 0.23–0.35 mg/kg/hour) but with comparable median values (0.29 vs 0.35 mg/kg/hour); ropivacaine infusion dosing was higher in the USA (median 0.5 mg/kg/hour, range 0.2–0.68 mg/kg/hour; n=392) compared with international dosing (median 0.2 mg/kg/hour, range 0.2–0.5 mg/kg/hour; n=129) (online supplemental table S9).

Table 2

Infusion dosing characteristics

Cumulative dosing (bolus plus infusion plus breakthrough)

While not usually reported in children, there are 24-hour dose limit recommendations in adults for ropivacaine13 and bupivacaine12 intended to reduce the overall risk of drug accumulation. As such, we report 24-hour cumulative doses (with the inclusion of initial bolus, infusion, and breakthrough dosing) as a potential risk factor for toxicity. Consistent with infusion rates providing the primary contributor to 24-hour dose, the dose of bupivacaine under 6 months of 8.3 mg/kg (range 3–13 mg/kg, n=91) was similar to the dose over 6 months of 8.4 mg/kg (range 3–25 mg/kg, n=333). Comparatively, the dose of ropivacaine under 6 months was 12 mg/kg (range 4.8–14.6 mg/kg, n=53), which was slightly lower than the dose above 6 months of 14.5 mg/kg (range 4.8–17 mg/kg, n=468). This occurred in the setting of infusions in neonates that either avoided a bolus31 51 or used a small bolus (eg, 0.5 mg/kg).38 TAP catheters represented a minority of catheters (48 out of 945 patients, 5%), and except for a case series of six cases,39 all used doses below the median 24-hour dose of PVB catheters.

Duration of infusion

Median duration of infusion was 72 hours (range 24–221 hours, n=945). Duration under 6 months was 43 hours (range 24–72, n=82) and 48 hours (24–120 hours, n=51) for bupivacaine and ropivacaine, respectively. Duration over 6 months was 48 hours (24–221, n=342) and 72 hours (24–192, n=470) for bupivacaine and ropivacaine, respectively.

LAST and toxic blood levels

We recorded a single case of LAST in a ropivacaine catheter, representing 0.2% of ropivacaine catheters based on naïve pooling (1 out of 521) and 0.1% of all catheters based on naïve pooling (1 our of 945). The toxic event occurred following a breakthrough bolus of choloroprocaine (table 3). This case was screened as a case report but not included due to the less than 24 hours of infusion.69 Three studies reported serum blood levels of bupivacaine above 2.1 mg/L, potentially toxic in humans (table 3).70 All patients underwent cardiac surgery. Two studies identified individual patient data (n=4) with bupivacaine levels of 2.81, 3.29, 3.14, and 3.21 mg/L.25 27 Another 10 infants were depicted graphically, with error bars extending above 2.1 mg/L, and received a bupivacaine infusion of 0.5 mg/kg/hour with a peak of 3.14 mg/L.5 All cases used infusions of 0.25–0.5 mg/kg/hour in 13 infants and 1 toddler. Peak concentrations occurred after 24 hours. Only total blood levels were monitored and not unbound serum levels.

Table 3

Cases of LAST and toxic blood levels


Herein we evaluated the dosing and toxicity of PVB and TAP catheters in infants and children. We assessed these blocks due to their historical character and large dose of local anesthetic typically used. The sites have divergent absorption kinetics but similar elimination half-lives71–74 that should predict risk of accumulation and toxicity over prolonged periods. Many patients tolerated doses of ropivacaine up to 0.5 mg/kg/hour for an average of 72 hours and doses of bupivacaine of up to 0.35–0.4 mg/kg/hour up to 48 hours with a low rate of toxicity. The only case occurred due to a 29 mg/kg breakthrough bolus of chloroprocaine, reinforcing package insert dosing recommendations.

The median hourly infusion dose of 0.33 mg/kg/hour for bupivacaine and 0.5 mg/kg/hour for ropivacaine reflects the potency equivalence of bupivacaine:ropivacaine of 1.0:1.5.75 However, based on median bolus doses of 1.25 mg/kg for bupivacaine, we would have anticipated a ropivacaine median bolus of 2 mg/kg instead of the 2.5 mg/kg reported. While reports from the USA represented a majority of cases, they were the minority of cases in patients less then 6 months, perhaps due to preference for chloroprocaine infusions in infants76 or use of epidurals.77 The majority of RCT data arose from Egypt and Iran in both infants and children representing a potential opportunity for multinational randomized study networks. Overall, randomized trial data in pediatric truncal blocks are sparse, with a recent meta-analysis of PVBs only including six single-shot trials.78 We were surprised by the paucity of TAP catheters with only 5% of all cases using TAPs; this contrasts with the adult data, where ~30% of adult catheters were TAP (Bungart et al. 22)


The median bolus dose of bupivacaine and ropivacaine comported overall with established dosing recommendations (eg, less than 2.5–3.0 mg/kg).16 79 However, maximum reported doses of 3.13 mg/kg for bupivacaine and 5.0 mg/kg for ropivacaine, all in older children, exceeded these limits.10 80 All infants under 6 months received bolus doses of less than 2.2 mg/kg of bupivacaine or ropivacaine.

Guidelines for infusion rates of bupivacaine and ropivacaine in children are less well established than bolus delivery and mostly based on published data from neuraxial infusions. Starting in 1992, several authors recommended infusion rates to minimize risk of toxicity,7 8 17 with textbooks advising bupivacaine limits of 0.2–0.25 mg/kg/hour for infants less than 6 months and up to 0.5 mg/kg/hour for older infants and children.15 With the introduction of ropivacaine, academic pediatric institutions raised the hourly limit to 0.6 mg/kg/hour (author’s clinical experience). In 2018, guidelines for dosing peripheral nerve catheters advised 0.2% ropivacaine or bupivacaine with rates of 0.1–0.3 mg/kg/hour irrespective of age10 but lacking a clear rationale.11 Neonates and younger infants are dosed differently due to increased local anesthetic accumulation and subsequent toxicity as a result of immature liver metabolism and lower levels of plasma-binding proteins, leading to longer half-lives of both bupivacaine and ropivacaine.80 Consistent with the textbook recommendation, 93% of infusions in patients greater than 6 months were run at 0.5 mg/kg/hour or less. However, 39% of patients less than 6 months received infusions greater than 0.3 mg/kg/hour, exceeding published dose limits. This dosing contrasts the PRAN analysis of 307 neonates, where only 20% received doses of greater than 0.3 mg/kg/hour of ropivacaine or 0.2 mg/kg/hour of bupivacaine in neuraxial catheters.77 The elevated neonate dosing primarily reflects data from Europe and UK in the 1990s5 24 and a single paper from Iran in 2010,31 but also includes multiple cases from the USA in 201139 and 2019.64 It illustrates the importance of international workgroups10 and global communication of risk factors to improve patient safety.


Complications including toxicity were inconsistently reported with an obvious lack of reporting standard for these outcomes. A single case of toxicity was observed at a rate (0.1%–0.2%) consistent with retrospective studies81 and the accompanying adult data (Bungart et al.22)Of note, the toxic case was the result of a bolus of chloroprocaine and not due the continuous infusion.48 69 Five cases of potentially toxic blood levels occurred, all in infants undergoing cardiac surgery before the year 2000. These reports arose in case reports/series and observational studies where LAST was not characterized in a systematic fashion. The small numbers of patients studied and a certain amount of chance likely prevented manifestations of LAST in the youngest infants. Infants cannot report symptoms of LAST, and it is possible that non-cardiac LAST was under-reported.21 82 Conversely, alpha-1-acid glycoprotein is upregulated following surgery, and total levels may rise while unbound levels stay in a safe range.83

We compared our data to cases from the major reviews of LAST in the literature characterizing data from 1979 to 2009,84 2010–2014,85 2014–201686 and 2017–2020,87 and a pediatric specific review of data from 2014 to 2019.82 In those reviews, we identified an additional nine catheters in pediatric patients. Four early cases of epidurals used high infusion or bolus values including 0.5 mg/kg/hour,88 1.25 mg/kg/hour,88 2.5 mg/kg/hour,89 and 7.5 mg/kg as divided boluses in 4.5 hours.89 Two cases used actual body weight instead of ideal or lean body weight with 0.25–0.35 mg/kg/hour90 split between multiple catheters in a 15 year old and a 13 year old. The remaining patients had risk factors including an intravascular catheter,91 carnitine deficiency due to prolonged valproic acid use,91 hypoalbuminemia,92 elevated transaminases,92 and breakthrough bolus doses.69 Cumulative risk factors are listed in box 1.

Box 1

Risk factors for local anesthetic systemic toxicity identified in catheters

Young age (lack of CYP3A4).

High weight-based dose.

  • Multiple catheters.

  • Use of actual body weight in calculations.

  • Breakthrough doses.


Cardiac surgery.

Liver dysfunction/resection.

Metabolic disease (carnitine deficiency from valproic acid use).

Comparison with adult population

The adult population had similar characteristics (eg, unilateral PVB in thoracic surgery) but with the single caveat of more TAP catheters (~30% compared with 5%). Many TAP catheters in adults were used in obstetrics (which does not apply to pediatric patients) and urological procedures (where caudal or penile blocks may predominate). In regard to dosing, adult populations have recommended 24-hour upper-limit doses of 400 mg for bupivacaine12 and 770 mg for ropivacaine,13 which are absent in children, except for infusion limits (of note, the package insert advises a limit of 28 mg/hour of ropivacaine in adults, leading to 0.4 mg/kg/hour in a 70 kg patient). Compared with the accompanying adult data (accompanying paper: Bungart et al. 22) infusion rates and 24-hour doses were lower in adults but with similar toxicity rates (0.15%–0.4%). The most common symptoms in adults were mild central nervous system features including confusion, agitation, somnolence, and dysgeusia that may go unreported in pediatric populations. Second, the larger sample size provided by adult publications may identify more rare events. Finally, pediatric anesthesiologists may provide better dose restrictions in the setting of ‘at-risk’ patients (eg, neonates, liver disease, hypoalbuminemia, and cardiac surgery) or through the use of chloroprocaine.93 Both adults and pediatric patients had similar risk factors with high doses as a major contributor. Given the overlapping data, pediatric dosing recommendations may provide a template to improve dosing in sick adult populations.


There are certain limitations to our analysis. In particular, we limited our results to the inclusion of catheters dosed for more than 24 hours and thus did not include two seminal papers from Lönnqvist6 94 with dose greater than 0.6 mg/kg/hour. Our weight-based dosing calculations used reported weight, not appropriately adjusted for lean body dosing. Our assumptions about case duration (2.5 hours) and patient weights may impact the results, but this only applied to four patients in total (0.4% of all catheters). Our characterization of LAST and toxic blood levels may overestimate or underestimate true rates due to low numbers. The published pediatric literature concerning pharmacokinetics of local anesthetics is based predominantly on infusions run during neuraxial blocks on which we base much of our discussion. Lastly, the printed literature may not reflect actual clinical practice as demonstrated by divergence in our results from previous PRAN papers77 95 96; thus, it may negate the generalizability of these findings (in the USA).

Conclusions and remaining questions

In conclusion, we investigated dosing of truncal peripheral nerve catheters based on the published literature of PVB and TAP catheters. Bolus doses frequently comported with dose limits with occasional doses above conventional limits.80 High infusion dosing was observed in both infants under and over 6 months of age despite age-related changes in local anesthetic half-life, which are inversely proportional to age. Both children under 6 months and over 6 months tolerated infusions up to 0.5 mg/kg/hour without symptomatic toxicity. However, toxic blood levels were observed in infants under 5 months receiving bupivacaine at 0.25 mg/kg/hour.25 Practitioners must remain vigilant about calculating dose limits in children of all ages but especially young infants. Given the paucity of safety events, rates of both bupivacaine and ropivacaine above 0.3 mg/kg/hour may be tolerated well in the appropriate older age groups. While we do not specifically discuss efficacy, we know that prior metaregression has identified that higher doses are more effective than lower doses for pain control.97 Following the ‘wisdom of crowds’ effect,98 we can speculate about effective and safe dosing using the interquartile and upper limit data. The extracted data would suggest the following doses, which mostly comport with the ASRA/ESRA international workgroup dosing data for epidurals10:

  • Ropivacaine bolus : 2.0–2.5 mg/kg using 0.2 or 0.5% with maximum of 2.5–3.0 mg/kg.

  • Ropivacaine infusion: 0.3–0.5 mg/kg/hour using 0.1–0.2% with maximum of 0.5 mg/kg/hour.

    • Reduced in 0–4 months to max of 0.2 mg/kg/hour and 5–12 months to 0.3 mg/kg/hour based on prior safety studies.7 17

  • Bupivacaine bolus: 1.2–1.4 mg/kg using 0.25% with max of 2.2–2.5 mg/kg.

  • Bupivacaine infusion: 0.25–0.35 mg/kg/hour using 0.1% with max of 0.5 mg/kg/hour.

    • Reduced in 0–4 months to max of 0.2 mg/kg/hour and 5–12 months to 0.3 mg/kg/hour based on prior safety studies.

  • Duration: 48–72 hours, maximum 120 hours.

Of note, pain control can often be achieved with doses much below the limits. Further research is also needed. Specifically, the field would benefit from guidance about ropivacaine and bupivacaine individualized dosing by age and block location, breakthrough doses from nursing, use of continuous infusion versus intermittent bolus methods, total of 24-hour dose limits, standardized reporting of complications in the pediatric literature, and recommendation about the use of chloroprocaine as an alternative in younger patients. A follow-up systematic review with multiple databases and meta-analysis or metaregression to assess for the contribution to toxicity or toxic blood levels in both pediatric and adult datasets would be worthwhile. Additional clarity will enhance the care of infants and children.

Ethics statements

Patient consent for publication

Ethics approval

Not applicable.


Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • Twitter @mfettiplace, @BrittaniBungart

  • Contributors MF conceived the project and designed the extraction. BB, LJ and MF contributed to literature screening, data extraction, data preparation, manuscript drafting and manuscript editing. KB contributed to data interpretation, manuscript drafting and manuscript editing.

  • Funding MF was supported by an National Institutes of Health T32 training grant (grant number 5T32GM007592-42).

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.

Linked Articles