Delphi Survey of International Pharmacology Experts: An Attempt to Derive International Recommendations for Use of Medicine in Breastfeeding Women

Abstract

Introduction:

There are currently no common guidelines used by health professionals to aid decision-making around the use of medicines during breastfeeding. Several specialized books, Web sites, and drug information services exist; however, all use slightly different criteria to derive their “safety hierarchy.” The aim of this study was to build consensus among international experts in pharmacology and breastfeeding to develop an agreed-upon classification system for safety of medicine use in breastfeeding women.

Study Method:

A three-round Web-based Delphi qualitative research design was used.

Results:

Seventeen experts in pharmacy/pharmacology and breastfeeding identified 15 key parameters that are used to inform decisions about medicines and breastfeeding. The most important parameters about the infant were the age and health of the child, and those of the medicine were the safety profile and experience of use in infants. The experts had a clear understanding of the complexity of decision-making when prescribing for breastfeeding women. Although the current number or letter classification systems do not incorporate important considerations such as infant age, a longer “descriptive text” incorporating all considerations may be impractical in busy clinical practice.

Conclusions:

Although clinicians and lay people would appreciate a simple classification scheme, in practice, decision-making about the safety of medicines for breastfeeding women is complex.

Introduction

Maternal medicines rarely cause adverse effects in breastfeeding infants.¹ With the exception of illicit drugs, radioactive substances, and antineoplastic medicines, medications can usually be prescribed to a breastfeeding woman while she continues to breastfeed.² Temporary cessation of breastfeeding or a “pump and dump” approach is associated with significant, and often underappreciated, adverse effects for maternal and infant health, including the removal of maternal antibodies, hormones, and fatty acids from the infant's diet, disruption of maternal milk supply, disruption of the mother–infant relationship, and potential risks associated with replacement feeds such as contamination of infant formula with bacteria or environmental toxins.³ However, many professionals continue to recommend temporary or complete cessation of breastfeeding or avoidance of medication “to be on the safe side.”

Historically, health professionals have received little formal education on breastfeeding medicine and medication issues. In addition, those searching for information on breastfeeding and safety of medicines are met with a confusing array of information sources.⁴ There are currently no common guidelines used by health professionals to aid decision-making around the use of medicines during breastfeeding. Several specialized books, Web sites, and drug information services exist; however, all use slightly different criteria to derive their “safety hierarchy.”

Many of the current guidelines use a categorical system for stratifying risk and providing recommendations. The publication Medications and Mother's Milk⁵ is well known among lactation experts and uses a five-tiered system (Table 1). In the past, the American Academy of Pediatrics used three categories (“compatible,” “may be of concern,” and “significant effects”),⁶ but more recently it has avoided individual guidance for each medicine, instead referring to LactMed, which does not feature a ratings system.^2,7 LactMed is a free online resource from the National Library of Medicine; it provides a summary at the top of each monograph.⁷ Medication Safety in Pregnancy & Breastfeeding⁸ and Drugs During Pregnancy and Lactation⁹ also avoid categories and provide short monographs synthesizing information on the medicine to facilitate individual interpretation of the evidence. The Australian Perinatal Psychotropic Medicine Information Service uses two categories: “caution” and “contraindicated.”¹⁰ Also in Australia, the Royal Women's Hospital¹¹ and the Australian Medicines Handbook¹² each have their own classification system. Not only are there many different categorization systems, but each system uses slightly different criteria for determining a medicines safety profile. Akus and Bartick⁴ evaluated 10 resources used in the United States for breastfeeding and medicines; of the 14 medications surveyed, there was no agreement between all resources on a medication's safety, and many resources failed to include the most up-to-date information in their analysis.

Table 1.

Classification Scheme Used by Hale ⁵

Classification	Description
L1 (safest)	Drug that has been taken by a large number of women without observed adverse effects in the infant. Controlled studies in breastfeeding women fail to demonstrate a risk to the infant, and the possibility of harm to the breastfeeding infant is remote, or the product in not orally bioavailable in an infant.
L2 (safer)	Drug that has been studied in a limited number of breastfeeding women without an increase in adverse effects in the infant and/or the evidence of a demonstrated risk that is likely to follow use of the medication in a breastfeeding woman is remote.
L3 (moderately safe)	There are no controlled studies in breastfeeding women; however, the risk of untoward effects to a breastfed infant is possible, or controlled studies show only minimal nonthreatening adverse effects. Drugs should only be given if the potential benefit justifies the potential risk to the infant. (New medications that have absolutely no published data are automatically categorized in this category, regardless of how safe they may be.)
L4 (possibly hazardous)	There is positive evidence of risk to a breastfed infant or to breastmilk production, but the benefits from use in breastfeeding mothers may be acceptable despite the risk to the infant (e.g., if the drug is needed in a in life-threatening situation or for a serious disease for which safer drugs cannot be used or are ineffective).
L5 (contradicated)	Studies in breastfeeding mothers have demonstrated that there is a significant and documented risk to the infant based on human experience, or it is a medication that has a high risk of causing significant damage to an infant. The risk of using of the drug in breastfeeding women clearly outweighs any possible benefit from breastfeeding. The drug is contraindicated in women who are breastfeeding an infant.

Medication safety in breastfeeding is ideally evaluated via clinical evidence of maternal milk levels, infant plasma concentration, and observed clinical effects. However, such studies are often unfeasible. Instead, recommendations may be based on pharmacokinetic parameters of the medicine, such as its molecular size, protein binding capacity, and bioavailability.¹³ Interpretation of this evidence is not straightforward, and the weight given to clinical versus pharmacokinetic factors may ultimately influence whether a medication is labeled safe by a particular resource. Consequently a medicine classified as low risk in one resource may receive a higher risk categorization in another.

Decision-making for health professionals is further complicated by pharmaceutical companies' reluctance to provide recommendations without a clinical evidence base. The Food and Drug Administration (FDA) have announced changes to their medicine labeling requirements, requiring pharmaceutical companies to provide a risk summary for both pregnancy and lactation.¹⁴ The current A, B, C, D, and X pregnancy categories will be removed in favor of a monograph detailing the potential risks to milk production and the infant, clinical considerations to consider when prescribing, and an overview of data on which the recommendations are based.¹⁴

It is unclear what information should be communicated to health professionals and women to aid decision-making for breastfeeding and medicines. The aim of this study was to build consensus among international experts in pharmacology and breastfeeding to develop an agreed-upon classification system for safety of medicine use in breastfeeding women.

Materials and Methods

A three-round Web-based Delphi qualitative research design was used. The key feature of the Delphi technique is the recruitment of an expert panel, which is then systemically surveyed about a specific question or issue.^15,16 Initial statements are derived from background research and analysis of the literature. Panel members provide researchers with individual responses to statements that are then compiled, summarized, and fed back to the group. The process repeats until consensus (a predetermined level of agreement) is reached. The Delphi technique offers advantages over traditional consensus building (e.g., face-to-face meetings) in that it promotes anonymity within the group, is cost-effective, limits geographical barriers, and supports reflective responses.^17–20

Ethics approval for this study was obtained from the Faculty of Health Science Human Research Ethics Committee at La Trobe University, Melbourne, VIC, Australia (protocol number FHEC 13/193).

Panel selection

Experts were identified based on a history of publications on breastfeeding, lactation, and medications and/or international prominence in breastfeeding and medicine education and training. Sixteen international experts were initially identified. In August 2013, an invitation letter was sent via e-mail providing a summary of the study design, its objectives, and the anticipated time commitment. Experts were also asked to nominate colleagues they felt would positively contribute to the study. This snowball recruitment method ensured greater diversity among the panel. Twelve additional experts were invited to participate based on panel recommendations. A positive response to the invitation letter served as informed consent.

Delphi process

In all three rounds, questionnaires were distributed electronically by a secure, password-protected, Web-based software program designed for creating and publishing surveys (Qualtrics, Provo, UT). Questionnaires were initially pretested on academics with expertise in pharmacology and lactation who were not participating in the study. Overall, the questionnaires balanced the needs of building consensus with allowing innovative ideas to surface in the early stages of the study. Experts were given 3 weeks to respond to each survey. Questionnaires were emailed on Day 0 of each round, and reminder e-mails were sent at Days 7, 10, and 14 for nonresponders. Nonresponders from a round were eligible for inclusion in subsequent rounds of the survey. Data were analyzed using descriptive statistics. Mean ranks, range, and percentage agreement were obtained for each parameter using the Qualtrics statistical program and Microsoft^® (Redmond, WA) Excel^®.

Round 1

Round 1 included several open-ended questions exploring current medicines and breastfeeding classification systems and their strengths and weaknesses. Participants were asked to name and describe the current classification systems for safety of medicine use in breastfeeding that they “know and/or use regularly.” Participants were then asked exploratory questions regarding important maternal, infant, and medicine parameters they consider when deciding on the safety and compatibility of a medicine in breastfeeding. Information on the proposed FDA pregnancy and breastfeeding labeling system (as described above) was also presented for opinion. The Round 1 questionnaire was sent in September 2013.

Responses obtained in Round 1 were summarized. Maternal, infant, and pharmacology parameters were standardized and formated into a 5-point Likert rating scale. Two authors (L.H.A. and C.B.) performed the qualitative thematic analysis. Parameters were tabulated in a Microsoft Excel spreadsheet and grouped according to their meaning; like statements (e.g., health of infant, comorbidities of baby, and any health problems in baby) were condensed into a single term. Following refinement, 31 parameters were fed back to participants in Round 2.

Round 2

Participants were asked to rate the degree of importance they place on the 31 infant, mother, and medicine parameters obtained from responses in Round 1 using a 5-point Likert scale (from 1=unimportant to 5=very important). Parameters were divided into three categories (infant parameters, mother parameters, and medicine parameters), and experts were asked to rank the three most important parameters in each category.

Participants were provided suggested modifications to the criteria in Hale's Medication and Mother's Milk⁵ and asked their agreement with proposed changes: “Do you agree with the following statements related to expanding the L3 ‘Moderately Safe’ category to include descriptive subcategories? (e.g., L3a—new medication no studies available; however, medicine likely to be safe based on pharmacokinetic profile; L3b—controlled studies show minimal nonthreatening adverse effects in infant; L3c—no controlled studies in breastfeeding; risk to infant unknown).” A 5-point Likert scale was used (1=strongly disagree; 2=disagree; 3=neither agree nor disagree; 4=agree; 5=strongly agree). Participants were also invited to explain their reasoning and propose additional modifications.

We also suggested a risk calculator and asked participants' opinions. One possibility to improve the current decision-making capacity of clinicians would be to introduce a risk calculator, allowing for decisions to be tailored to the individual infant, the mother, and the medicine being prescribed. An example of such a risk calculator currently being used is the New Zealand Cardiovascular Risk Calculator (see below). Further details can be found on the National Prescribing Service Web site.²¹ A medicine and breastfeeding risk calculator would be based on categories relative to the infant, mother, and medicine (e.g., gestational age, relative infant dose, or medicine molecular weight). Participants were asked to agree/disagree with statements about the use of a risk calculator (5-point Likert scale) and to select up to four options as being the most important factors, labeling them 1 to 4.

Round 3

Round 3 sought to obtain consensus on parameters explored in Round 2 through rating and ranking techniques. The 15 most important infant, mother, and medicine parameters from Round 2 were fed back to the panel to allow for convergence of opinions on key parameters.

Results

Sixteen experts were initially invited to participate in the study, of whom 12 expressed an interest in being involved. Ten additional experts were sent invitations based on snowball recommendations, of whom five participated in the study. The final panel was composed of 17 experts based in Australia, the United Kingdom, the United States, and Norway (listed in Acknowledgments, with permission). Participants came from either a pharmacy or pharmacology background, with the majority having a master or doctorate degree in addition to their primary tertiary qualifications. All experts were working in the field of breastfeeding and pharmacotherapy, including primary research, advisory services, and teaching of health professionals. Responses were received from 13 participants for Round 1, 14 for Round 2, and 12 for Round 3.

Contemporary classification systems

Round 1 explored contemporary classification systems and the key advantages and disadvantages of current systems. Medicines and Mother's Milk⁵ was the most commonly listed classification system (mentioned by 13 participants). One participant mentioned the Australian publications, Therapeutic Guidelines,²² which use “a simple four category classification that ranges from compatible to caution for a drug about which we have little data but whose effects are of little concern to avoid….The fourth category is avoid based on high transfer to milk and concerning side effects.” One participant mentioned the Swedish categories,²³ adding, “I am not sure if it was ever implemented. It has all the disadvantages of any classification system.”

LactMed, Briggs' Drugs in Pregnancy and Lactation, and the American Academy of Pediatrics were also mentioned, although they do not provide a classification system. Independence from the pharmaceutical industry was seen as a benefit of both LactMed and Medicines and Mother's Milk.⁵ The simplicity of numerical risk categories had a mixed response, with some experts believing it enabled healthcare practitioners to provide advice: “with a quick glance at the L risk number, clinicians can have an overall idea of suitability.” Another said, “Simple, easy to follow, authoritative.” Others believed that the categories could lead to uninformed decision-making as it “leads to a cessation of thought about the topic—just quote the category—does not take infant age into account [and] can lead to [poor] advice.” Other comments included, “Some users think only L1 drugs can be taken while breastfeeding,” and “L3 and L4 are gray areas.” One participant explained, “L3 particularly can be confusing. A drug can be placed in L3 because of completely lacking excretion data but be pharmacologically unable to penetrate milk to any degree. The L3 [rating] probably implies a greater risk than is necessary. Small gripe for what I consider the best available.”

Classification systems were further explored in Round 2, with the majority of experts (9/14) believing that separating hazards associated with breastmilk production and hazards that directly affect the infant would be beneficial.

Key parameters

In Round 1, participants listed the maternal, infant, and medicine parameters that they considered important (Table 2). Participants rated these in Rounds 2 and 3, as described in Materials and Methods.

Table 2.

Parameters Listed in Round 1

Parameter	Number of participants mentioning (n=13)
Maternal and infant characteristics
Age of infant	13
Health of infant	10
Overall balance between the benefits of breastfeeding and advantages of medicine versus the risk of formula and potential effects of medicine	5
Number of breastfeeds per day/volume of milk	5
Maternal need for medicine	4
Monitoring for infant	3
Gestational age	2
Infant renal and hepatic function	2
Mother's health	2
Mother's wishes regarding breastfeeding	2
Infant weight	1
Strategies available to reduce exposure of infant to medicine	1
Risk associated with stopping breastfeeding	1
Exposure to medicine in pregnancy	1
Medicine characteristics
Safety profile of medicine	13
Relationship between clinical dose and medicine in breastmilk	6
Alternative medicines available	5
Experience of medicine use in infants	5
Dose of medicine required	4
Absolute/relative infant dose	3
Pharmacokinetics	3
Duration of treatment required	2
Route and timing of medicine administration	2
Gastrointestinal absorption of medicine	2
Postmarket experience with medicine	2
Quality of and quantity of data available on medicine	2
Molecular weight of medicine	1
Protein binding capacity of medicine	1
Effect of medicine on lactation	1
Time to peak blood/milk level	1
Half-life of medicine	1
Passage of medicine across infant blood–brain barrier	1
Infant blood level	1
Polypharmacy	1

In response to the question,“What do you think are the most important parameters for health professionals and women to consider when deciding on the safety and compatibility of medicine use in breastfeeding?”

Based on respondent ratings, 15 key infant, maternal, and medicine characteristics were identified as being important in making decisions about medicines and breastfeeding (Table 3). Consensus was reached on the three most important parameters in the decision-making process. The “overall balance between the benefits of breastfeeding and advantages of medicine versus the risks of formula and the potential adverse effects of medicine” was considered the most important parameter, followed by the age of the infant and the safety profile of the medicine. These three parameters had the highest rankings in rounds 2 and 3 and had mean Likert scores of >4.40.

Table 3.

Infant, Maternal, and Medicine Parameters Important for the Decision-Making Process

Rank	Parameter	Mean rank	Standard deviation	Mean Likert score
1	Overall balance between the benefits of breastfeeding and advantages of medicine versus the risks of formula and the potential adverse effects of medicine	1.7	1.89	4.83
2	Age of infant	3.8	1.93	4.75
3	Safety profile of medicine	4.7	2.54	4.42
4	Experience of medicine use in infants	5.5	3.50	3.75
5	Health of infant	5.9	3.57	3.83
6	Relationship between clinical dose and medicine in breastmilk	7	2.58	4.00
7	Mother's wishes regarding breastfeeding	8.1	5.11	4.25
8	Number of breastfeeds per day/volume of milk consumed per day	8.8	2.35	3.92
9	Half-life of medicine	8.9	3.07	3.92
10	Monitoring of infant available	9.3	2.75	3.83
11	Gastrointestinal absorption of medicine	10.1	3.63	4.17
12	Alternative medicines available	10.7	1.77	3.75
13	Protein binding capacity of medicine	10.9	4.65	3.33
14	Quality and quantity of data available on medicine	11	4.22	4.00
15	Strategies available to reduce exposure of infant to medicine	13.6	1.35	3.00

As “monitoring of infant available” was important to some respondents, in Round 2 we asked, “What does this mean to you?” Respondents explained that they meant clinical observations as well as testing where applicable. One response shows the detail involved:

A huge range of parameters which MAY include ongoing regular visual assessment of alertness, growth, energy, disposition, muscle tone, colour, sucking reflex, sleep patterns, etc, regular assessment by independent health care personnel, appropriate blood, urine, milk tests (only when appropriate and necessary), appropriate communication options for questions/concerns including links to quality, peer-reviewed information and advice, knowledge of ‘warning signs’ and appropriate responses to them.

Proposed FDA labeling

Experts were asked their opinion on the FDA recommendations, as well as their impact, utilization, and practical application. There was universal agreement that the new descriptive labeling was an improvement on the current system. However, many experts were wary of the practicality of such a system, citing pharmaceutical medicolegal considerations and the time constraints of clinicians as reasons for the proposed medication monographs failure: “it [the FDA medicine monograph] looks like a significant improvement… unfortunately medico-legal considerations may trump clinical experience hence very conservative recommendations” and “personally I think some of these labels will be over long, and most doctors will not read it all” were typical comments.

Risk calculator

The respondents mostly agreed that “The risk calculator would improve how risk is communicated to clinicians” (11/13). However, most “neither agreed nor disagreed” with the other statements about the use of a risk calculator. Comments ranged from positive, such as

A well designed risk calculator should be a feasible option, if it is user friendly. A sample isn't provided, so can't comment on whether this one might be “too complicated.” Presuming the calculator factors in individual characteristics, it absolutely should provide an individual mother/baby risk assessment.

to more negative:

The decision on breastfeeding while taking medications is a multifactorial individualised decision. I don't think a risk assessment such as that above can cover all the factors.

and

The whole survey seems premised on the use of one medication at a time. My impression is that many if not most adverse reactions occur in infants whose mothers are taking more than one medication, particularly CNS [central nervous system] depressant drugs, at the same time.

Discussion

Main findings

This study identified 15 key parameters that are used by experts to inform decisions about medicines and breastfeeding (Table 3). The highest ranking item was about risk versus benefit, so is not actually a parameter. The most important parameters about the infant were the age and health of the child, and those of the medicine were the safety profile and experience of use in infants. Qualitative feedback from participants suggests that prescribing for breastfeeding women is a nuanced process that requires careful consideration and that a single “one size, fits all” classification system might lead to uninformed and incorrect decision-making by clinicians.

Strengths and limitations

To our knowledge this is the first attempt to explore expert opinions on the decision-making process involved in medicines for breastfeeding women. Our panel included key decision-makers in this field, including the authors of several of the key texts on the topic. The study was conducted via Web-based communication and maintained anonymity among participants. The demographic characteristics of the group indicate that participants had the required skills and experience to make informed contributions to the discussion.

A three-round Delphi approach was undertaken. This allowed experts to complete the study in a timely manner without developing survey “fatigue.” The three rounds allowed us to formulate a list of key parameters that inform breastfeeding and medicine use decisions, but further work is needed to determine how these parameters might be used by clinicians when prescribing medicines to mothers. In particular, we explored the idea of constructing a “risk calculator” similar to those used for cardiovascular risk scores.²¹ A risk-model approach was seen by experts as an improvement on the current classification systems; however, the complexity of constructing such a model was beyond the scope of this study.

A limitation of this study was that the panel was composed mostly of experts from English-speaking and developed countries. This may restrict the applicability of our findings to other countries; however, we believe that the decision-making criteria for medicines and breastfeeding are universal, and although the medical conditions and medications may change between countries, the principles underlying the decision-making process should stay the same.

Study implications

The importance of breastfeeding is well established; however, clinicians are still reluctant to prescribe medicines during breastfeeding, leading to many women temporarily or permanently stopping breastfeeding during illness or not taking beneficial medicines while breastfeeding. Establishing a clear set of internationally recognized parameters to inform the decision-making process could be the first step to improving breastfeeding mothers' access to, and experience with, medicines.

Our results show that pharmacology experts have a clear understanding of the complexity of decision-making when prescribing for breastfeeding women. It remains a challenge to translate this complexity into clear messages that can be accessed by time-poor clinicians and are meaningful for women and their families. Although the current number or letter classification systems do not incorporate important considerations such as infant age, a longer “descriptive text” incorporating all considerations may be impractical in busy clinical practice. More work is needed on establishing a balance between these two scenarios. This study provides a starting point for such work.

Conclusions

Although clinicians and lay people would appreciate a simple classification scheme, in practice, decision-making about the safety of medicines for breastfeeding women is complex. One participant summarised the issues in this way:

There needs to be a greater education of the public and medical profession about the mechanics of breastfeeding (it is not possible for a mother to stop for 2 days, and then just start up again) and the health and economic benefits of breastfeeding. The message needs to be that hardly any drugs are a problem for breastfeeding mums, and that it is almost never necessary to stop breastfeeding because of maternal medications. We need a good easy-to-understand information source that mothers and health care professionals can access 24/7.

Footnotes

Acknowledgments

This study would not have been possible without the experts who contributed their time to the process. We would like to acknowledge Phillip Anderson, Cheston Berlin, Delwyn Cupitt, Tom Hale, Ken Ilett, Majella Hill, Wendy Jones, Sue Jordan, Debra Kennedy, Judith Kristensen, Yuan Loke, Treasure McGuire, Frank Nice, Hedvig Nordeng, Hilary Rowe, Tricia Taylor, and Rodney Whyte for their contribution. Funding was received from the Faculty of Health Sciences, La Trobe University (2013).

Disclosure Statement

No competing financial interests exist.

References

Anderson

, Pochop

, Manoguerra

. Adverse drug reactions in breastfed infants: less than imagined. Clin Pediatr (Phila), 2003; 42:325–340.

Sachs HC; Committee on Drugs. The transfer of drugs and therapeutics into human breast milk: An update on selected topics. Pediatrics, 2013; 132:e796–e809.

McDonald

, Amir

, Davey

M-A

. Maternal bodies and medicines: A commentary on risk and decision-making of pregnant and breastfeeding women and health professionals. BMC Public Health, 2011; 11(Suppl 5):S5.

Akus

, Bartick

. Lactation safety recommendations and reliability compared in 10 medication resources. Ann Pharmacother, 2007; 41:1352–1360.

Hale

. Medications and Mothers' Milk, 15th ed. Hale Publishing LP, Amarillo, TX, 2012.

American Academy of Pediatrics Committee on Drugs. Transfer of drugs and other chemicals into human milk. Pediatrics, 2001; 108:776–789.

National Library of Medicine. Drugs and Lactation Database (LactMed). Available at http://toxnet.nlm.nih.gov/cgi-bin/sis/htmlgen?LACT (accessed March 24, 2009).

Koren

. Medication Safety in Pregnancy & Breastfeeding: The Evidence-Based A-to-Z Clinician's Pocket Guide. McGraw Hill, New York, 2007.

Schaefer

, Peters

, Miller

. Drugs During Pregnancy and Lactation: Treatment Options and Risk Assessment, 2nd ed. Elsevier, London, 2007.

10.

Royal Women's Hospital. Perinatal Psychotropic Medicine Information Service. Available at www.ppmis.org.au/ (accessed February 26, 2015).

11.

Loke

. Pregnancy and Breastfeeding Medicines Guide. Pharmacy Department, The Royal Women's Hospital, Melbourne, Australia, 2010.

12.

Rossi

. Australian Medicines Handbook 2012. Australian Medicines Handbook Pty Ltd., Adelaide, 2012.

13.

Rowe

, Baker

, Hale

. Maternal medication, drug use, and breastfeeding. Pediatr Clin North Am, 2013; 60:275–294.

14.

Federal Register. Available at http://federalregister.gov/a/2014-28241 (accessed January 7, 2015).

15.

Pill

. The Delphi method: Substance, context, a critique and an annotated bibliography. Socioecon Plann Sci, 1971; 5:57–71.

16.

Powell

. The Delphi technique: Myths and realities. J Adv Nurs, 2003; 41:376–382.

17.

Fiore

Jr , Bialocerkowski

, Browning

, et al. Criteria to determine readiness for hospital discharge following colorectal surgery: An international consensus using the Delphi technique. Dis Colon Rectum, 2012; 55:416–423.

18.

Holloway

. Doing the E-Delphi: Using online survey tools. Comput Inform Nurs, 2012; 30:347–350.

19.

Jones

, Hunter

. Consensus methods for medical and health services research. BMJ, 1995; 311:376–380.

20.

Lisby

, Nielsen

, Brock

, et al. How should medication errors be defined? Development and test of a definition. Scand J Public Health, 2012; 40:203–210.

21.

National Prescribing Service. New Zealand Cardiovascular Risk Calculator. Available at www.sld.cu/galerias/pdf/servicios/hta/ebm_cardio_new_zeland.pdf (accessed June 6, 2014).

22.

Therapeutic Guidelines. Available at www.tg.org.au/ (accessed February 26, 2015).

23.

Sannerstedt

, Berglund

, Flodh

, et al. Medication during pregnancy and breast-feeding—A new Swedish system for classifying drugs. Int J Clin Pharmacol Ther Toxicol, 1980; 18:45–49.