Pill Counting as an Intervention to Enhance Compliance and Reduce Adverse Outcomes with Analgesics Prescribed for Chronic Pain Conditions: A Systematic Review

Abstract

Purpose of Review

Appropriate use of opioid analgesics is a key concern within the field of pain medicine. Several methods exist to discourage abuse and facilitate effective treatment regimens. Pill counting is often cited as one such method and frequently employed in varying fashions within clinical practice. However, to date, there is no published review of the evidence to support this practice. This was a comprehensive review of the available literature that was conducted with analysis of the efficacy and practical application of pill counting during treatment of chronic pain conditions.

Recent Findings

There is paucity in data regarding pill count importance in pain management. Pill count is a very important tool to monitor compliance of opioids use which in turn can prevent several complications associated with opioid misuse.

Summary

Pill counting may be used in conjunction with other abuse deterrents, although increased support for this practice requires standardized methods of pill counting and further analysis of its effectiveness.

Introduction

The risks of opioid overdose and abuse continue to be preeminent concerns in the field of pain medicine. Although those committed to eradicating opioid analgesia from the pain management armamentarium intentionally conflate prescription opioid overdose deaths with those due to illicit fentalogues and recent data indicating that involuntary tapers of opioids actually increase the likelihood of death from overdose and suicide, the need for continued diligent focus on prescription practices by clinicians on behalf of their patient population remains a best practice [1, 2]. Multiple strategies have been proposed to mitigate the risks of inappropriate medication use both wittingly and unwittingly. Methods today may include urine drug testing (UDT), pain diaries, patient contracts or agreements, governmental prescription monitoring programs, electronic container monitors, and ingestible markers [3, 4]. Pill counting is often mentioned in studies on medication effectiveness or adherence with the goal of verifying the actual medication intake pattern. This method of medication monitoring has been implemented with pharmacotherapies for multiple pathologies, including lipid-lowering, HIV treatment, contraception, depression, and sickle cell disease [5–9]. Pill counting may involve a variety of approaches, including the requirement of the patient to present their current medication supply for verification at the clinic, at their home, at a local pharmacy, or via telephone or video calls over scheduled or variable intervals. It is also often mentioned in guidelines aimed at the prevention of opioid abuse in chronic cancer and nonmalignant pain [10, 11]. This review seeks to clarify the literature regarding pill counting as an effective and practical method to enhance compliance with analgesics prescribed for chronic pain conditions.

Methods

This study reviewed the literature to appraise the rationale and efficacy of pill counting in the management of outpatient oral analgesic regimens. To maintain transparent and detailed standards, Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were followed throughout the search and review. Searches were performed of the electronic databases PubMed, Scopus, MEDLINE, and the Cochrane Database for manuscripts indexed from the inception of each database until September 2021. The search terms included “pill count*,” filtered by “opioid,” “narcotic,” “chronic,” and “analgesic” in each respective database. Identified studies were uploaded into Mendeley, and duplicates were removed. Abstracts were screened to identify meta-analyses, randomized controlled trials (RCTs), case–control and cohort studies, quasi-experimental studies, observational studies, case series, and case reports. Selection criteria were applied by two independent reviewers, and discrepancies were resolved by a third author. Inclusion criteria focused on English-language human studies that investigated the efficacy of pill counting as a risk mitigation tool in chronic pain conditions. This included outcomes such as misuse of opioids or clinical pain score changes, objective functional results, and/or depression and anxiety scales. To investigate the efficacy of pill counting as an adherence monitoring instrument, studies were excluded if there was non-medical opioid use (i.e., use of opioid medication without a prescription or use with a prescription but not as prescribed). Studies were also excluded if pill counting was not used as an outcome measure or was a non-principal component of the study (Fig. 1).

Fig. 1.

Flow chart of article search and inclusion process

Results

A total of 82,718 manuscripts were identified in the initial search, and 2277 duplicates were removed. After filtering by keyword, 336 articles remained. Titles and abstracts were reviewed for pill counting as a component of chronic pain treatment. Of the 24 identified full-text manuscript with potential eligibility, 20 lacked a focus on pill counting efficacy, and one study used medication that was not prescribed as an analgesic. Three articles met inclusion and exclusion criteria and were reviewed (Table 1). All studies were prospective in nature using multicomponent interventions including controlled substance agreements, random drug screening through UDT, and pill counts. No study exclusively isolated pill counts as an intervention but rather used it as a component of the study interventions. Outcome measures included opioid pill confiscations by law enforcement and proportion of patients with aberrant drug-related behaviors [12–14].

Table 1.

Analysis of articles meeting inclusion criteria

Study	Focus	Design/methodology	Sample size	Participant characteristics	Condition	Analgesic	Study duration
Bujold et al. [12]	To describe the effects of practice guidelines on reducing the diversion of prescription drugs	Prospective cohort study	Participants: n = 27 (primary care clinicians)	Primary care clinicians treating patients in a rural setting	Chronic nonmalignant pain	Non-specific opioids	24 months
Brown et al. [13]	To evaluate divergent drug-related behaviors among patients with chronic pain and investigator compliance with universal precautions	Open-label, nonrandomized, non-comparative, multicenter, prospective	Participants: n = 1487 (patients), n = 281 (investigators)	Mean age of patients: 53, 57% female, 92% with chronic pain > 1 year	Chronic, moderate-to-severe pain ≥ 3 months	Morphine sulfate-extended release	Not clear
Manchikanti et al. [14]	To evaluate controlled substance abuse after implementation of a monitoring program	Prospective with historical controls	Participants: n = 500	Mean age of patients: 48.5, 59% female; duration of pain mean: 10.7 years	Chronic pain	Hydrocodone, oxycodone, methadone, morphine	Not clear

Bujold et al. studied the implementation of practice guidelines intended for primary care clinicians in a rural community and their effects on opioid pill confiscations by law enforcement [12]. The guidelines encouraged patient-signed pain contracts, use of random UDT, and random pill counts. A survey was sent to 35 primary care clinicians with a subsequent 77% response rate: 90% indicated that they used the guidelines developed by the task force. Over a 2-year period, confiscation of opioid pills decreased by 300%.

Brown et al. analyzed the feasibility and compliance of a universal precautions (UP) approach to determine the risk of aberrant drug-related behavior and to guide management of patients with chronic, moderate-to-severe pain being treated with morphine sulfate [13]. Components of the UP approach included treatment agreements, tracking of prescriptions, use of the Screener and Opioid Assessment for Patients with Pain^®-Revised questionnaire, pill counts, pain patient follow-up tools, investigator assessment/plan, and urine drug screens. Using the UP approach, 1487 patients in a primary care setting were assigned a risk level for opioid abuse/misuse at each of the five visits with their care provider. Following the completion of the study questionnaire, 80% of the investigators who responded opined that pill counts were a useful or very useful utility and 58% indicated that they would continue to utilize pill counts in their practices. Aberrant drug-related behavior was detected throughout the study, which supported the feasibility of a UP approach for assessing aberrant drug-related behavior.

Manchikanti et al. aimed to identify controlled substance abuse following the implementation of a controlled substance agreement [14]. Five hundred patients who were receiving stable doses of hydrocodone, oxycodone, methadone, or morphine were followed prospectively. All participants signed consent and controlled substance agreements. This agreement allowed the investigators to review charts and collect information regarding controlled substance intake, which included random drug screening, pill counts, and education. Overall, 9% of participants were determined to be abusing prescription drugs. When compared to historical controls, this demonstrated a 50% reduction in opioid abuse.

Discussion

The optimal use of opioid analgesics remains a crucial focus of pain management principles. The emphasis on a clinician’s duty to reduce abuse raises the question of optimal prescription monitoring. Despite pill counting being frequently mentioned throughout the literature as a method for promoting adherence to pain medication regimens, it is poorly defined, sparsely studied, and inconsistently implemented. This systematic review serves to determine the utility of pill counting as an intervention to improve compliance and safety in patients who are prescribed opioid medication for chronic pain conditions. In addition, this review demonstrates that the prescribing clinician’s compliance tools and methods are respectfully utilized to the fullest and assist him/her in efforts to utilize this risk mitigation tool. Based on the review’s data, the clinician will ideally become a safer and thus more effective prescriber through pill counting. Following extensive analysis, there is a dearth of high-quality studies that focus on pill counting as an intervention. There were no studies that exclusively isolated pill counting as an intervention despite the existence of many studies in which the authors posit on the perceived utility of pill counting. These studies, however, did so without actual supportive data. Rather than being incorporated as an intervention, pill counting was included as a component in a monitoring program or practice guideline, although without analysis of efficacy.

Several studies mention the utility of pill counting as a component of opioid management; however, few include specific citations for these statements. Of the 24 full-text manuscripts assessed with potential eligibility and not included in the final analysis, only nine specifically reference the utility of pill counting [15–23]. Colasanti et al. cited an article by Chou and colleagues, in which the authors state, “Because patient self-report may be unreliable for determining amount of opioid use, functionality, or aberrant drug-related behaviors, pill counts…can be useful supplements” (p. 123) [15, 24]. Likewise, Viscomi et al. cited guidelines by Trescot and colleagues, in which the authors state, “Adherence monitoring is crucial to avoid abuse of the drugs and at the same time to encourage appropriate use, and involves the initiation of drug screening, pill counts, and patient care agreements” (p. 41) [16, 25]. In general, these statements regarding the utility of pill counting in opioid monitoring appear to be consensus statements and author opinions rather than empirically derived. These articles do not refer to specific analyses or trials of pill counting utility in opioid management.

The simple count of current pill supply may be a more sensitive and objective measure than other methods to detect aberrant behavior for purposes of therapeutic adjustment or signaling warnings. Logical reasoning suggests that an inappropriately low supply could be related to abusive medication patterns or inadequate coverage of pain at a given dosage. However, it also stands to reason that patients may forget to bring medications for verification, discard unused doses (“pill dumping”), or share medication with other individuals [3]. Excessively high supply may signal memory problems in adhering to a regimen or the need to decrease prescription quantities. Logical reasoning affirms this method’s limitation outside of discrete formulations such as capsules or tablets. Furthermore, the pro re nata (PRN) use of medications will result in a more variable representation [26]. Since pill counting is inherently calculated on the quantity of dispensed medications, it is not able to account for ingestion of medication from non-prescription or other sources, such as previous surplus, those obtained from another provider, or those obtained through diversion.

Ultimately, feasible implementation of pill counting, in and by itself, seems to have limited utility aside from representing a potentially problematic snapshot of current medication stock. If the concern is patient disuse of medication, excess supply may be artificially decreased by simply not providing all available medications or disposing of unused pills. There is no guaranteed relationship between the number of pills remaining and those actually ingested by the patient. If the concern is patient excessive use of medication, pill counting is again fraught with issues of supply inputs or medication obtained from additional sources and overall remains an unproven system for promoting safe medication practices. The resources expended by the physician and clinical staff to enforce pill counting provides additional concern for the implementation of this measure. If technology is leveraged to assist with the process, such as a mobile telephone application which would allow for instantaneous verification of current pill supply via photo on the device, then it may provide greater utility for tracking medication use and providing clinicians with additional data necessary to adjust prescriptions. Furthermore, placing the verification in the patient’s hands may attenuate feelings of distrust between the provider and patient. All of this may occur with pill counting implemented in the clinic setting.

Regarding limitations of this review, the most notable one is the limited number of studies that met the inclusion criteria limiting the strength of the conclusions that we have drawn. Hopefully, this review will inspire more high-quality prospective studies on the efficacy of pill counts as a risk mitigation tool and consequently increase the strength of the support for pill counting in risk mitigation platforms.

On a final note regarding ethical issues, social media is becoming rife with claims and copies of actual opioid agreements, indicating that certain prescribers are requiring patients to appear at their physicians’ offices within 2 h of notification in order to undergo a random pill count. As a result, patients in these practices are “handcuffed” by this requirement and are not allowed to travel whatsoever due to the risk of not being able to show up for the pill count within the given time frame. Particularly, given that many patients were unable to visit family members who lived out of state for more than 2 years during the throes of COVID-19, such draconian policies are highly unethical, representing a clear assault on the bioethical principle of respect for patient autonomy. Any physicians who are engaged in such problematic practices need to reconsider their practices of “chaining” their patients to the practices physically.

Conclusion

The clinical utility of pill counting remains limited due to feasibility concerns as presented above. Although frequently suggested as a component of a monitoring program or practice guideline, there is limited evidence to support the use of pill counting as a stand-alone approach to risk mitigation. Pill counting in the form in which it has been utilized thus far may be helpful when used in conjunction with other methods such as random urine drug screens. There is a great need for more feasible and precise methods for pill counting to improve the validity and reliability of its utilization. Engaging in more sophisticated pill counting approaches supports the notion that practicing clinicians are doing all that they can to treat their pain patients in the safest, most effective ways available to them, which equates to improved overall quality in the practice of pain medicine.

Compliance with Ethical Standards

Conflict of Interest

The authors declare no competing interests.