Chemicals from the practice of healthcare: Challenges and unknowns posed by residues in the environment


  • Published on the Web 8/14/2009.

The practice of health care often relies heavily on the use of a bewildering array of chemicals for diagnostics, therapy, prophylaxis, and lifestyle or cosmetic modification. Excretion, bathing, manufacturing, and disposal of pharmaceuticals and personal care products (PPCPs) serve as conduits to the environment for complex mixtures of parent chemicals and transformation products, primarily via sewage and domestic refuse. As members of a much larger universe of natural products and other anthropogenic chemicals that already pervade the environment, PPCPs enter the environment primarily from multitudes of individually miniscule sources. Each source by itself contributes relatively insignificant quantities, but the combined inputs can yield measurable levels in waters and other environmental compartments, with the general exception of air. Scenarios abound for chronic, low-level ambient exposure of wildlife, microbiota, and humans, but special situations can lead to higher level, acute exposures. Whatever the existing risks, they can span a wide spectrum of modalities and can be difficult to decipher because of the complexities posed by simultaneous exposures to numerous chemical stressors—perhaps each present below any individual level known to perturb biological processes—and some leading to difficult-to-detect or delayed-onset subtle effects.

The study of PPCPs in the environment (PiE) has proved challenging over the course of the last 15 to 20 years of international research. Significantly, the ultimate aims of PiE research are sometimes unclear. Overall priorities need to be established to achieve outcomes that still remain to be articulated. While the published scientific literature has grown to thousands of papers targeted primarily at deciphering the shape, scale, intensity, and spatiotemporal aspects of the environmental footprint, exposure envelope, and potential for biological effects of PPCPs, many aspects of PiE remain obscure. Given the possible reality of continually diminishing resources for research, a concerted effort is needed to identify those select aspects capable of removing the most uncertainty in assessing whatever risks might be posed by PiE, target those aspects having the highest potential to broadly benefit human health and the environment, and better coordinate and focus future research.

The study of PiE is notable in that it requires expertise spanning a remarkably diverse spectrum of disciplines—ranging from hydrology, civil engineering, and chemistry, to pharmacology, toxicology, medicine, and even social psychology and risk communication. Study of PiE has captured the attention of not just scientists, but also policy makers, legislators, regulators, environmental agencies, health-care communities, public, press, and the pharmaceutical, pharmacy, and health insurance industries. It has also slowly morphed into the much larger issue of the so-called but loosely defined “emerging contaminants”—a catch-all term for contaminants whose presence or significance was previously unknown, unrecognized, or underappreciated [1].

Why does PiE persist as a topic of interest for so many? A major reason is that despite the accelerating pace of published investigations, new questions continue to be generated while some major ones remain unanswered. Moreover, the fact that PPCPs (a term coined 10 years ago [2]) occur in waters serves to illustrate the intimate connections between the activities and behaviors of humans and the environment—a continual reminder of the hydraulic connectivity between sewage and “natural” waters. They remind each of us that we are integral parts of the water cycle. Their seeming ubiquity in waters, especially potable waters, serves as a constant (and sometimes emotional) reminder that these waters originated at least in part from the excretions of others—from feces, urine, and sweat [3]. This factor certainly looms large with the life cycle of PPCPs and can play a critical role in the public acceptance of water reuse.

Pharmaceuticals and personal care products comprise thousands of distinct chemical entities and tens of thousands of commercially formulated products, possessing an immense range of physicochemical and physiological properties. For drugs, each active pharmaceutical ingredient (API) can be assigned to one of many therapeutic groups, such as those of the tiered Anatomic Therapeutic Chemical (ATC) classification system or the analogous system for veterinary medicines (e.g., see discussion in Table 5 of Ruhoy and Daughton [4]). Because of the extreme diversity of PPCPs (especially their wide range of biochemical activities), generalizations applied across the entire spectrum (or even within defined classes) are prone to misrepresentation and numerous exceptions. Even broad classes sharing the same therapeutic modalities (e.g., lipid regulators [ATC C10] or antidepressants [ATC N06A]) can act via a wide variety of biochemical routes. Despite these categories of therapeutic action, actual mechanisms of action are frequently unknown and individual APIs can be promiscuous in their biochemical actions.

Among the reference citations in the U.S. Environmental Protection Agency's (U.S. EPA) bibliographic database on PPCPs (, publications with a focus on personal care products comprise a much smaller portion (about 10% of the total) than do APIs. The major personal care product groups that have been investigated are the synthetic musks, triclosan and triclocarban, ultraviolet filters and sunscreens, parabens, and siloxanes, in decreasing order of prevalence in the literature. Phthalates, bisphenol A, and nonylphenols are involved with higher numbers of publications but their usage in personal care is minor compared with other commercial uses. In general, the active agents in the thousands of commercial formulations of personal care products are produced and consumed in much larger quantities than APIs but their biological potencies are also much lower.

Nearly all aspects of APIs in the environment have parallels for both human and animal pharmaceuticals; indeed, many APIs have dual uses. The relative importance, however, of these aspects among human and animal applications can differ greatly because of the dominance of (and special needs imposed by) confined animal feeding operations in the overall use of APIs targeted for animals, where antibiotics and the endogenous and synthetic steroids play dominant roles. The focus of roughly 10% of the articles inventoried in the U.S. EPA bibliographic database on PPCPs is veterinary and aquaculture usage. Perspectives on the roles of veterinary medicines as environmental contaminants have been covered in a number of excellent reviews, including those published in Crane et al. [5].


Environmental scientists and health-care professionals face nearly endless challenges with PPCPs. But which of these challenges leads to valuable near-and long-term outcomes that protect and improve human health and ecological function? We must also determine where PPCPs fall within the growing list of overarching environmental issues in a world of diminishing resources and continually emerging new concerns.

The issues and concerns surrounding PiE involve the interface between humans and the environment—where the everyday individual actions, activities, and behaviors of multitudes of people combine and intersect with the environment via dynamic transfer and recycling of countless different chemicals, most of which were designed to impart biological effects. Targeting of research will require integration of knowledge regarding the presence, fate, and effects of PPCPs in the environment with what is known about the countless sources and origins of their release as a direct result of the management and administration of health care. A key insight into this challenge is that a wide spectrum of actions targeted at reducing the transfer of PPCPs to the environment holds the potential for reciprocally improving the quality and costs of health care [6]. Treating the environment and health care as an integral system could greatly clarify where and how to invest resources to achieve optimal outcomes, as improvements in either can lead to collateral improvements in the other.

Using prioritization tools such as multi-criteria decision analysis coupled with value of information, and examining the continuum of steps spanning the risk paradigm—beginning with sources and origins and ending with biological effects and risk management—could be used to establish relative priorities; for examples of this process, see Kiker et al. [7] and Linkov et al. [8]. Nearly every stakeholder involved with PiE serves not only as an interested party, but also as an actual contributor to some aspect of the overall problem, as well as a potential beneficiary from solutions. Each stakeholder also can play an active role as problem solver; the physician can alter prescribing habits, the consumer can make more prudent purchases and properly dispose of leftover drugs, and the insurer can encourage dispensing of prudent quantities. To identify the relative importance of each modification needed to improve a system as large and complex as health care requires establishing clear priorities, which in turn must be based on quality data.


Of the 7,000 or so references currently captured in the U.S. EPA bibliographic database on PPCPs (over 85% of which are articles from journals or books but largely omitting the non-English literature), over 90% have been published only since 1999 (Supporting Information, Fig. S1;–138.S1). The international extent of the topic is evident from the numbers of publications that feature a particular country in the abstract or title. Over 1,600 articles mention 10 different countries (Australia, Britain, Canada, China, Europe, Germany, Italy, Japan, Spain, and Sweden) and Europe, and over 600 mention the United States. Of 150 academic dissertations, 60% are from outside the United States. Of the seven most highly cited papers on PiE, five originated in Europe and two from the United States.

While this certainly shows an ongoing escalation in publishing activity, it does not tell us if these works have targeted the most pressing needs, if they are being actively used to inform decision making, or whether they are resulting in useful outcomes for society. Moreover, if one were to assume an extremely conservative cost of merely US$10,000 to $100,000 per paper, the last 10 years of research targeted toward PPCPs may have consumed minimum resources roughly upwards of US$600 million. Such a substantial investment prompts the question of whether these resources could have had more productive outcomes or greater impact if they had been invested elsewhere in the field of PPCPs or, perhaps more significantly, even elsewhere in environmental sciences at large. Perhaps not surprisingly, the impressive wealth of data published on the topic of PiE has generated a host of new questions, which, paradoxically, can serve to breed yet more uncertainty. At the same time, however, the new knowledge gained for PPCPs is often directly relevant to other types of chemical contaminants, serving to leverage resources throughout the environmental sciences arena.

Two overarching concerns for PiE have centered on human health risk and ecological integrity, especially aquatic effects from the perpetual entry of residues via sewage (pseudo-persistence). The ultimate destination for PiE research might be evident only in the larger context involving a truly holistic examination of PiE and the complete life cycles of PPCPs. Can health-care systems and the manufacturing and distribution of PPCPs be designed and optimized to leave a minimal environmental footprint? The argument has been made that in taking actions to reduce and ultimately minimize an ill-defined hazard (namely, the release of PPCPs to the environment) by reinventing health-care administration and delivery of health care, improvements in therapeutic outcomes might follow naturally, together with reductions in the costs associated with medical care [6].

The best outcomes regarding release to the environment might emerge from optimizing the way in which health care and personal care are administered, distributed, prescribed, dispensed, and employed and how PPCPs are designed and produced. Reactive approaches using end-of-chain controls are not as efficient or effective as proactive solutions that optimize source reduction or pollution prevention. By focusing on less sustainable end-of-pipe solutions, such as improved ways to dispose of unwanted medications or more efficient treatment of wastewaters, approaches with even better outcomes might escape consideration. As one example, consider that the imprudent usage of PPCPs coupled with the extent of leftover medications can be viewed as direct measures of the inefficiencies and wastefulness that can occur along the entire life cycle of PPCPs. Leftover medications represent much more than just chemical wastes needing disposal. They represent wasted health-care resources, inflated and unnecessary consumer expense, and missed opportunities for achieving optimal therapeutic outcomes [6]. By focusing on controlling the many causes leading to the accumulation of unwanted medications, not only would the need for disposal be reduced, but excretion and discharge of residues of PPCPs might also be incidentally reduced as a result of optimized usage. Such holistic approaches require the involvement of specialists from fields that may not have foreseen ever playing active roles in the PiE issue.


In the early 1920s, Henry Ford conceived of a new strategy for inventory maintenance designed to improve return on investment. Called “just-in-time,” this new paradigm redefined on-hand inventory as essentially being the equivalent of waste. Optimal performance meant perfect balance between demand and on-hand supply. If a just-in-time perspective were applied to health care, medication waste could be viewed not just as additional chemical contaminant burden for the environment, but more importantly as a prime metric of inefficient, nonoptimal administration of health care. Redesign of health care using the just-in-time perspective and the knowledge and expertise of medical practitioners, health-care administrators, pharmaceutical manufacturers, and environmental scientists could lead to a holistic system of balanced and optimally targeted delivery of medical care. Such a system could yield improved therapeutic outcomes, lowered costs, and reduced environmental impact.

The effectiveness of efforts directed at pollution prevention or source reduction increase as the targeted steps reside closer to the source or origin of the chemical. Tracing the ultimate origin back to chemical design, advancements in eco-design could prove to have significant outcomes not only in reducing environmental impact, but also in improving health-care outcomes ([6,9];

In the near term, consideration could be given to the design of pilot projects designed around stewardship actions in health care. Healthcare organizations having control over all aspects of medical care might serve as excellent testing grounds for pilot projects; in the United States, one example for testing new approaches could be the nation's largest integrated healthcare system—the Veterans Health Administration.


Despite the thousands of publications devoted to the many facets of PiE, unanswered questions persist and continue to proliferate. Many of these questions, however, are also germane to some of the major issues that permeate environmental science as a whole rather than being critical to solving specific problems associated solely with PiE. Significantly, despite the wealth of published data, little has yet proved of use in actual implementation of system redesigns that are more sustainable or even for informing regulatory deliberations regarding PiE.

A major weakness in the application of environmental science to PiE has been the failure to frame the issue in a much larger context—using a “systems” approach that involves experts from fields other than primarily just analytical chemistry and environmental engineering. A comprehensive, international strategy for tackling PiE using a harmonized approach integrated across a spectrum of disciplines could also involve social psychologists and risk communicators, physicians, pharmacologists, pharmacists, drug designers, and health insurers.

Once the PiE issue is successfully framed in a larger, holistic context having meaning to a broader audience, and collaborations are established among those from across disparate disciplines, more productive outcomes could possibly emerge. The requisite framing must show how health care and personal care can lead directly to environmental contamination. But more importantly, clear communication would be essential for how measures directed at redesigning their administration to minimize the PiE footprint can in turn improve the affordability and desired outcomes from the consumer use of PPCPs. Required actions could become clearer and more readily embraced when considering the patient and the environment as an interconnected whole.

Although voluminous published and gray literatures exist for PiE, a large number of gaps remain that could be investigated [10]. Science is never short on questions. More important to address first, however, is what exactly do we wish to accomplish with more research? What outcomes are we seeking? How would the uncertainty associated with assessing risk be most efficiently minimized? Closer collaboration between researchers and risk assessors would be highly beneficial. Most importantly, however, the concerns, challenges, and solutions regarding PiE need to be framed and examined in the expanded context of the larger systems used for the care of human and animal health.

In the table (Supporting Information, Table S1;–138.S2), some notable gaps and unresolved questions are compiled, and examples are provided to illustrate a range of issues that might be considered for further attention and discussion. The following introduces these questions and concerns.

Inter connectedness and unintended consequences

Solutions that might seem at first to solve a certain problem can have unintended, and sometimes adverse, consequences. The interconnectedness of our world might seem obvious from our vantage point today, and it is certainly embodied in the new “systems” disciplines. But the realization that “everything is connected to everything else” was first formalized less than 40 years ago (in 1971) by Barry Commoner as his “First Law of Ecology.” Garrett Hardin later reformulated the idea in “we can never do merely one thing” (Hardin's Law). That unanticipated or unforeseeable outcomes can result from a single action was captured by what Crawford Holling later called “environmental surprise,” where the ultimate outcome can differ dramatically from what was anticipated. But the forerunner to the modern environmental movement was George P. Marsh, who, more than 150 years ago, recognized that interconnections pervade all of nature: “No atom can be disturbed in place, or undergo any change of temperature, of electrical state, or other material condition, without affecting, by attraction or repulsion or other communication, the surrounding atoms. These, again, by the same law, transmit the influence to other atoms, and the impulse thus given extends through the whole material universe” [11]. Some examples of unanticipated consequences pertinent to PiE are provided in Table S1.

Mining the published PiE literature

Multi-criteria decision analysis and value of information could serve as the principal means for setting PiE priorities. A number of areas on face value seem to deserve concerted attention. Many have been delineated in publications and various government reports. A primary need is to capitalize on what is already available—the published literature, which could contain a wealth of data not yet examined and certainly never thoroughly mined, compiled, summarized, evaluated, and distilled into useful insights and knowledge. This is largely because synoptic reviews and compilation of prior data are generally not valued in science as much as the publication of new data. But given the uncertainty as to the extent these published findings have been evaluated (evidence exists that much of the published literature in general has never even been read [12]), the value of publishing without clear outcomes in mind must be questioned. This prompts the more general questions of how the impact of publishing can be improved, and how do we encourage the capture and synthesis of this hidden knowledge?

An examination of the U.S. EPA bibliographic database on PPCPs reveals that publishing on the topic of PiE began in earnest around 1996, which saw roughly twice as many articles as in 1995 (80 vs 40). The very first publications devoted specifically to the topic of PiE, however, began to appear in the 1970s. One of the very first significant works came from Tabak and Bunch at the U.S. Department of the Interior [13] followed six years later by Coats et al. [14]. The topic began to attract more than a thousand publications per year beginning in 2007. The first two months of 2009 showed more publications (256) than in all of 1999 (207); in the first 6 months of 2009, over 800 documents have either been published or are in press. A rapidly inflating literature (but not necessarily expanding in scope) greatly increases the possibility that an ever greater share of these publications will not receive adequate examination, as no longer can a single individual commit sufficient time to being thoroughly familiar with the literature as a whole; specialization in individual aspects of the topic is necessitated, and this can slow advancements in the absence of well-targeted cross-disciplinary collaborations. The lack of sufficient synoptic review greatly increases the risk of duplication of prior work [15] and in not focusing new work where the highest priority gaps might be. Some additional aspects of literature mining are summarized (Supporting Information, Table S1).

Notable gaps

Cursory examinations of the studies published to date hint that the majority of data appears to focus on environmental occurrence and monitoring and on treatability efficiencies for wastes and drinking water. At the other end of the spectrum significant areas have received surprisingly little attention. Some of the major ones are summarized (Supporting Information, Table S1).

An important aspect of environmental monitoring or site characterization studies rarely discussed is the veracity of structural identification of contaminant unknowns. The frequency with which PPCPs purportedly identified in environmental samples have undergone structural confirmation is unknown. This potential weakness leads to a number of questions (Supporting Information, Table S1).


Bringing to bear ever-more advanced measurement methods, analytical chemists allow us to peer into the shadows of chemical space with greater magnification and clarity. While this newly discovered chemical landscape might be fascinating to explore and serves to further illuminate the expansive universe of chemical stressor exposure, at the same time it poses greater challenges for risk assessors. This is especially true with regard to one of the greatest problems facing toxicology today—simultaneous/sequential chronic exposure to multiple chemical stressors, each present at ever-lower individual concentrations. While baseline narcosis is believed to be the most common mode of action at very low stressor concentrations, the possibility of unique, unpredicted mechanisms of action cannot be ruled out, especially since mechanisms of action can change with exposure levels (multi-phasic dose-response) and receptors can vary across taxa. Increasingly lower detection limits will pose greater challenges for assessing, communicating, and ameliorating diminishing risks. And an inflating known universe of potential chemical stressors will challenge the feasibility or sustainability of regulatory/compliance monitoring on a chemical-by-chemical basis.

At the other extreme, technological prowess in measuring the very small sometimes distracts or overshadows the potential for unanticipated scenarios for overt toxicity caused by acute exposures and poisonings—not just unintended poisonings of humans and pets from leftover medications, but also poisoning of wildlife via previously unrecognized source and exposure pathways and even unrecognized or unappreciated mechanisms of action. Can improved assessment of risk for PPCPs account for the possibility of unanticipated exposure scenarios or adverse outcomes? What additional knowledge is required to avoid the scale and consequences of acute-exposure incidents such as the mass poisonings of raptors and scavengers by pentobarbital or residues of nonsteroidal anti-inflammatories such as diclofenac (and possibly quinolone antibiotics) remaining in carcasses from medicated domestic animals, or to predict the unexpected acute toxicity of APIs to certain nontarget species?

Our environment extends beyond the confines of water, soil, and air. It also encompasses where we live and even our bodies themselves, where residues of countless chemicals are applied to or excreted from the skin and then transferred to other surfaces where others can then unknowingly be exposed [3]. Other scenarios for acute exposures therefore include interhuman contact (e.g., from high levels of APIs remaining on the skin after dermal application or excreted via the skin) and human contact with excretions from medicated pets; these issues are particularly germane for therapeutic treatments using highly cytotoxic or hormonal drugs.

With respect to prioritizing APIs for in-depth study, little evidence supports an overriding importance of toxicological data derived from therapeutic doses or commercial production volumes or usage rates. Also needing consideration are other critical variables involved with the life cycles of APIs, such as pharmacokinetics (e.g., the extent to which an API is extensively excreted unchanged—or as metabolically reversible conjugates—via feces, urine, or sweat), delivery route, patient compliance (lower rates of compliance generate leftovers and the consequent need for disposal), potency, usage patterns leading to episodic releases, and an API's propensity for off-target promiscuity.

A major problem with studies of long-term, low-level ecological exposure is the ever-increasing challenge of deconvoluting the occurrence of what might first appear to be an adverse effect in a certain (sub)population from the effect's frequency of incidence as ambient background. The problem of tying effects to exposure is discussed further (Supporting Information, Table S1); this problem also underlies regulatory frameworks based on assessment of hazard for individual chemicals. Moreover, even if causality can be established, the next hurdle must be faced: determining if effects actually impart detrimental, irreversible changes at the organizational level of the population.

Other gaps and research needs can be summarized along the continuum of the risk paradigm, spanning the range from chemical stressor sources to biological effects and remediation [10]. With respect to the topic of drug stewardship, the unknowns surrounding sources and origins are particularly important (Supporting Information, Table S1).


As with any field of study, always present is a risk that opinions, beliefs, and biases can become codified as dogma. Vigilance is needed to evaluate the veracity of purported facts, especially when misinformation might be unwittingly used to inform decision making. Some statements regarding PiE may be based more on suppositions than facts and might benefit from more investigation (Supporting Information, Table S1).


A final point involves the key importance in establishing and communicating the full context in which PPCPs exist in the environment as potential stressors for biological systems. Their place in the larger universe of chemical stressors is essential to appreciate so that the public can develop a more accurate and meaningful perspective of chemical exposure in general and that optimally informed regulatory decisions can be formulated. Diminishing resources for research must be directed to the most significant contributors of environmental risk. Of course, the true picture of the relative toxicological importance of PPCPs can only be obtained by considering all hazards, including nonchemical forms of stress, such as electromagnetic, radiological, biological, physical, thermal, noise, and emotional, among many others. This presents an enormous challenge—one that cannot yet be satisfactorily addressed—but should at least be continually maintained in the background of discussion and debate to ensure steady progress in eventually acquiring the ability to assess risk in a truly holistic manner.

In the mean time, we can at the very least accept the issue of PiE in the environment as an opportunity—as a driver for improving the efficiency and efficacy of the responsible contributory systems. By involving the many professional communities engaged in health care, sustainable systems can be designed that could yield substantive savings in health-care resources, improved patient outcomes, and combined protection of human health and ecological integrity.


Fig. S1. Yearly publications relevant to PPCPs.

Found at DOI: 10.1897/09–138.S1 (137 KB PDF).

Table S1. Surprises, questions, and framing a bigger picture regarding PiE.

Found at DOI: 10.1897/09–138.S2 (38 KB PDF).