Medication Use Among Patients Reporting Xerostomia of an Academic Dental Clinic

Abstract

Background:

Global prevalence of xerostomia has been reported at 22% (range 0.01%-45%), negatively impacting oral health, nutrition intake, and quality of life. The causal relationship between xerostomia and medications remains uncertain but greater understanding could guide interventions.

Objective:

To describe the demographic characteristics and medication regimens in patients with xerostomia of an academic dental clinic.

Method:

This is a retrospective academic dental clinic record review from July 1, 2018 to October 27, 2020. Patient records were obtained from the University at Buffalo, School of Dental Medicine. Xerostomia status was determined via query of electronic health records and validated by manual review. Pharmacologic class and xerostomic potential of medications were identified by the Veterans Affairs Drug Classification System and drug compendia, respectively. Predictors of medication use were assessed using a multiple logistic regression model.

Results:

Of 37 403 examined records, 366 (0.98%) were identified as xerostomic. After excluding confounding factors (Sjogren’s and radiation), 275 of 317 patients received at least one xerostomic medication, majority were female (240, 66%) versus male (126, 34%). Mean ± (SD) age was 64.9 ± 15.11 years. A total of 208 (57%) patients were aged ≥65. The median number of total and xerostomic medications were 8 (interquartile range [IQR], 4-12) and 4 (IQR, 2-7), respectively. The 3 most prevalent xerostomic pharmacologic classes were antidepressants (131, 35%), gastric medications (101, 28%), and vitamin D (87, 24%).

Conclusion:

Despite observed prevalence of xerostomia lower than global prevalence, xerostomic medication burden for patients experiencing xerostomia was high. Pharmacist-led interprofessional collaborations should be investigated to reduce xerostomic burden.

Introduction

Saliva is a vital component of maintaining oral health. Saliva is a source of lubrication to aid in swallowing, speaking, and allowing the comfortable movement of removable dental appliances. Patients who do not produce enough saliva suffer from a condition called hyposalivation. Although similar, hyposalivation is not the same as xerostomia. Xerostomia is the subjective sensation of dry mouth and may not relate to the degree of saliva production by the patient. ¹

Xerostomia with low salivary flow has a tremendous impact on oral health. Individuals with this condition are at increased risk of caries, poor mucosal health, and food avoidance. ² Xerostomia alone is associated with difficulties in eating, swallowing, speaking, and poor self-perceived health, which can also impact daily life. ³ Patients with xerostomia can have bacterial accumulation along with carious and broken teeth. This can lead to aspiration pneumonia, a major source of mortality for institutionalized elderly. ⁴ The components in saliva provide buffering (which prevents acid erosion of the teeth), as well as IgG, IgA, and other immune molecules to provide a first-line defense of invading microorganisms in the second most dense biome in the body. ⁵

Furthermore, xerostomia is associated with medical conditions such as arterial hypertension, diabetes, and bone degenerative, cardiac, pulmonary, psychological, or digestive conditions. ³ Especially in the aging population, medications used for treating these disorders have been implicated in exacerbating xerostomia, which consequently contributes to the geriatric syndrome of malnutrition and a decrease in quality of life.^6,7 Older patients often experience a phenomenon known as “Rapid Oral Health Deterioration,” coined by Marchini et al. ⁸ Rapid oral health deterioration is a condition of relatively sudden breakdown of the dentition and oral function. It leads to loss of quality of life and surgical treatment with some morbidity or mortality. The causes for this condition are multiple but xerostomia plays a major role in its natural history.

The global prevalence of xerostomia has been reported at 22%, with medication use as the most common etiology.^9-11 Xerostomia is such a common side effect that saliva’s vital importance may sometimes be minimized or its absence viewed as a minor inconvenience by the prescriber. Although the risk of xerostomia with medication use is well known, the nature of what medications contribute most to xerostomia is not well understood. Previous work has quantified that number of medications correlates with xerostomia and what medications pose a risk of xerostomia.^7,12-16 While these reports help identify medications that can contribute to xerostomia, greater understanding is still needed to determine what degree of exposure to these medications is needed to cause xerostomia in the patient. These reports did not specifically describe the number of xerostomic medications patients were exposed to. In addition, a number of these studies were conducted outside of the United States, potentially limiting their applicability to this population.

Objectives

The primary aim of this study is to describe the demographic characteristics and medication regimens of patients with xerostomia of an academic dental clinic in the United States. A greater understanding of the medication profiles of patients with xerostomia could foster opportunities for pharmacist-led multidisciplinary interventions to reduce patients’ xerostomic burden and improve quality of life. This study was pharmacist-led and may provide a template for future research by pharmacists using data from academic dental electronic health records.

Methods

Data Source

Analyzed records for this cross-sectional analysis were extracted from the electronic health record (EHR) of the University at Buffalo, School of Dental Medicine clinic (SDM). Records from July 1, 2018 (date of initiation of the electronic health record at the institution) to October 27, 2020 were included. The SDM EHR contains patient-reported medical history, medication history, and records of treatment received at the SDM clinic. Medical history is patient-reported. The SDM EHR does not have direct access to health records outside of the clinic. Medication data for the SDM EHR are imported from an external prescription claims database. The dental provider will verify if the medication is currently being received by the patient before adding to the EHR. Medication records unavailable through the prescription claims database are manually entered by clinic personnel. Patient-identifiable information was anonymized. Each patient record was assigned a unique study-specific identifier. This project was approved by the University at Buffalo Institutional Review Board (Study #4600) and reported following STROBE guidelines.

Data Query

The final search query was completed in collaboration with the SDM software engineer responsible for data extraction. Identification of patients with xerostomia was accomplished through multiple strategies. An initial query of keywords associated with xerostomia (Table 1) identified records with an indicator of xerostomia or dry mouth in a progress note. Diagnosis codes, procedure codes, and answers to standard form questions regarding xerostomia status were also searched. Patient demographic information and a text excerpt of 100 characters from before and after the xerostomia keyword indicator in the patient record were extracted. Following identification of records, a multidisciplinary team consisting of 2 pharmacists, a student pharmacist, and dentist manually reviewed records to confirm xerostomia status. Following review of the extraction, study personnel determined there was not sufficient detail to differentiate between dry mouth, hyposalivation, and xerostomia. For example, “hyposalivation” appears in the progress note as a patient complaint but no objective measure of salivary flow was found. Alternatively, “xerostomia” appeared as a patient complaint, but no objective measures of xerostomia were identified. Clinic personnel often use the terms interchangeably. For the purposes of this study, they are considered the same condition.

Table 1.

Search Terms and Diagnosis Codes Used to Identify Patients With Xerostomia.

Search terms	“xero”
	“hyposaliv,” “hypo saliv,” “hypo-saliv”
	“drymouth,” “dry mouth,” “dry-mouth”
	“sjo”
	“salivary flow,” “saliva flow,” “saliva-flow”
	“absence of saliva,” “abs saliva,” “abs-saliv,” “absence of salivary”
	“lack of saliva,” “lacks saliva,” “lack-saliv,” “lack of salivary”
	“salivary gland,” “saliva gland,” “saliva-gland”
	“mouth dry”
	“zerost” a
Diagnosis codes/Description	494038/Xerostomia
	408957/Sjögren Disease (Sicca syndrome)
	693126/Post-irradiation xerostomia
	494040/Xerostomia
	312932/Sjogren’s syndrome (Sicca syndrome)
	494039/Xerostomia
	41267/Hyposecretion

^aLooking for an obvious misspelling attempt of “xerostomia.”

Following the determination of patient xerostomia status, medication data were extracted for those patients identified as having xerostomia. Records indicating a predisposing condition for xerostomia (Sjogren syndrome, history of radiation therapy for head, neck, or throat cancer) were identified and excluded from final analyses to prevent confounding. Only prescription and over-the-counter medications active on the date of the initial xerostomia indicator were included in the analysis. Medication records identified as vaccinations, medical supplies, herbals, supplements, or vitamins were excluded. Vitamin D was included as an exception to the vitamin exclusion criteria given an identified risk of xerostomia. ¹⁷ Review of extracted medication names for correct spelling and commercial availability was conducted. Once the student pharmacist completed the initial review of medication names, 2 pharmacists reviewed 10% of the medication names for quality assurance. All medications reviewed were found to be spelled correctly and commercially available. These pharmacists also provided a secondary review for medication names that were unclear based on student pharmacist review.

Medication Classification

To facilitate analysis, medications used by patients with xerostomia were linked to a Veteran Health Affairs (VHA) drug class via RxNorm Concept Unique Identifier (RX CUI). The VHA drug class was used because of public availability and previous application in research. ¹⁸ In cases where a medication was a combination of multiple drug products, the record was divided into the individual medication elements with drug class assigned to each individual element.

Xerostomic potential of medications was assigned based on a collection of tertiary and systematic reviews.^6,12-14,17 A drug was determined to be xerostomic if the cited resource identified the drug as having dry mouth or xerostomia as an adverse event.

Statistical Analysis

Age, gender, total number of unique medications, total number of unique xerostomic medications, and total number of unique drug classes were reported using descriptive statistics. The Mann-Whitney U test was used to compare total number of xerostomic medications of patients aged greater than 65 years with the total xerostomic cohort. Due to the retrospective cross-sectional nature of the study, a sample size/power calculation was not performed. The medication profiles of patients with xerostomia were analyzed using a multiple logistic regression model. This analysis can be used to assess the association between multiple risk factors and dichotomous outcomes. ¹⁹ The variables used in the model were age (≥65 and <65 years of age) and gender (male, female) as these factors have been identified as being associated with xerostomia. ²⁰ Variables were analyzed using backward elimination with minimal significance set at 0.2 to make inclusion in the final multiple logistic regression model producing odds ratios (ORs) and 95% confidence intervals (CIs) for each variable. Analyses were performed using SAS version 9.4 (SAS Institute, Cary, North Carolina), and all hypothesis testing was 2-sided with a significance set at P < 0.05.

Results

Of 37 403 records reviewed, 366 (0.98%) patients were identified as having xerostomia (Table 2). Among these 366 patients with xerostomia, 49 were found to have a predisposing xerostomic condition. Of these patients, 44 were identified as having Sjogren’s syndrome and 5 had a history of radiation therapy to the head, neck, or throat. Following exclusion of patients with predisposing conditions, most patients (275, 87%) received at least one xerostomic medication. Most of the total cohort identified with xerostomia were female (240, 66%) versus male (126, 34%) with a mean (SD) age of 64.9 ± 15.11 years.

Table 2.

Characteristics of Patients With Xerostomia.

	Male	Female	Total
Patients identified as having xerostomia, No. (%)	126 (34)	240 (66)	366
Age (mean, SD)	66.9 ± 15.34	63.9 ± 14.95	64.9 ± 15.11
Age ≥65, No. (%)	81 (22)	127 (35)	208 (57)
Age <65, No. (%)	45 (12)	113 (31)	158 (43)
Receiving xerostomic medications (total), No. (%)	107 (29)	205 (56	312 (85)
Receiving xerostomic medications (w/o S/R), No. (%)	100 (27)	175 (48	275 (75)
Sjogren’s (S) a , No. (%)	5 (1)	39 (11)	44 (12)
Radiation (R) a , No. (%)	2 (1)	3 (1)	5 (1)
Receiving xerostomic medications (S/R), No. (%)	7 (2)	30 (8)	37 (10)

^aRemoved during analysis to avoid confounding factors.

Analysis identified a median number of medications used of 9 (interquartile range [IQR], 5-13) in patients with xerostomia who were 65 years or older (Table 3). When compared with median use of medications compared with the entire cohort with xerostomia, 8 (IQR, 4-12), the difference was not found to be statistically significant (P = 0.28). The median number of xerostomic medication was 4 (IQR, 2-7). The 5 most prevalent xerostomic pharmacologic classes were antidepressants (118, 37%), gastric medications (89, 28%), vitamin D (78, 25%), beta-blockers (76, 24%), and opioids (76, 24%) (see the Appendix for a list of medications included in these classes). Upon regression analysis, antidepressant use was more likely in the age group ≥65 years, OR of 1.927 (95% CI: 1.208-3.073, P = 0.0059) compared with <65 years. In contrast, antispasmodic use was less likely in the age group ≥65 years compared with <65 years with an OR of 0.404 (95% CI: 0.184-0.888, P = 0.024). The use of vitamin D and opioid analgesics was correlated with age as well (Figure 1). When analyzing xerostomic medication use and gender, statistically significant ORs were identified in the use of antidepressants and vitamin D (Figure 2).

Table 3.

Medication Utilization of Patients With Xerostomia.

n (median, IQR)	Male	Female	Age ≥65	Age <65	Total	Prevalence, %
Total medications	8, 4-12	8, 4-12	9, 5-13	7, 3-11	8, 4-12	N/A
Medication classes	7, 4-11	8, 4-12	8, 5-12	6, 3-10	8, 4-11	N/A
Xerostomic medications	4, 2-7	4, 2-7	4, 2-7	4, 1.5-7	4, 2-7	N/A
VHA Drug Class Code and Class Name a
CN609, Antidepressant	35	83	56	62	118	37%
GA900, Gastric Medications	28	61	56	33	89	28%
VT509, Vitamin D	18	60	52	26	78	25%
CV100, Beta-Blockers	35	41	51	25	76	24%
CN101, Opioid Analgesics	21	55	36	40	76	24%
CN601, Tricyclic Antidepressant	4	15	9	10	19	6%
AH100-AH109, Antihistamine	9	23	17	15	32	10%
GU201, Antispasmodics	12	25	28	9	37	12%

Abbreviations: IQR, interquartile range; VHA, Veteran Health Affairs.

^aReported as number of patients.

Figure 1.

Multiple logistic regression of selected medication classes by age (≥65 versus <65).

Abbreviations: CI, confidence interval; OR, odds ratio.

Figure 2.

Multiple logistic regression of selected medication classes by gender (female versus male).

Abbreviations: CI, confidence interval; OR, odds ratio.

Discussion

This report highlights an important clinical area for dentists and pharmacists to explore as an opportunity for collaboration. More than 85% of patients with xerostomia were receiving at least one xerostomic medication. Polypharmacy is a known risk for xerostomia that is reinforced by these results.^10,16 A previous study examined the number of medications used from an academic dental clinic in relation to xerostomia status but did not provide names or drug classes of the medications. ⁷ Our study adds to the literature by examining specific medication classes and their use in patients with xerostomia. A thorough medication review by the pharmacist for a patient complaining of xerostomia could identify medications that could be discontinued or switched to an option with a lower incidence of xerostomia. Dental settings with integrated pharmacist-led services have shown improved prescribing practices and use of medication. ²¹ Pharmacists collaborating with a dentist could lead to improved management of xerostomia as well.

Pharmacist-led utilization of electronic medication records from an academic dental clinic in the United States is relatively novel. This was a pilot program for pharmacist-directed research at our institution. This study has developed a process for future studies while also raising awareness of this potential avenue for research. As an institution contributing data to the Big Mouth Dental Data repository, the methods used for our institutional-level study could translate to the larger data set, spurring additional investigations.

The project also promoted informatics as a research pathway. Promotion of informatics and information technology into oral health has been identified as an area of need. ²² The intent is to continue to foster this collaboration with oral health care providers to enhance the pharmacist role in this setting. This will create foundational evidence for the clinical translation of pharmacist services to optimize the health outcomes of patients with oral diseases.

Antidepressants, as defined by the VHA drug class, include serotonin and noradrenaline reuptake inhibitors, selective serotonin reuptake inhibitors, serotonin modulators, bupropion, and mirtazapine. ²³ Although tricyclic antidepressants were recognized as the worst offenders among all antidepressants, other medications within the class can still exhibit dry mouth side effects. ⁶ The broad classification used in this study might have contributed to this class being the most prevalent drug class among patients with xerostomia. Besides antidepressants, the most common medication classes among patients with xerostomia were gastric medications, vitamin D, beta-blockers, and opioids. A similar trend was observed in the nationally representative data from the National Health and Nutrition Examination Survey (NHANES). ²⁴ In NHANES, the use of beta-blockers, antidepressants, narcotic analgesics, and prescription proton-pump inhibitors was among the top 18 used drug classes and increased in prevalence from 1999-2000 to 2011-2012. These similarities between our clinic and national trends suggest that local trends are consistent with what is reported nationally.

Figures 1 and 2 identify what medications pharmacists could expect patients with xerostomia to be receiving. Greater knowledge of these medications and their alternatives can guide training and education of pharmacists and dentists managing xerostomia. As the most prevalent drug class seen, greater familiarity with antidepressants would be valuable. In addition to xerostomia, antidepressants have been implicated in dental implant failure. ²⁵ Clinicians should consider the impact that antidepressant xerostomia may have on implant success.

The gastric medication group was primarily composed of proton-pump inhibitors such as omeprazole. Although proton-pump inhibitors do not possess anticholinergic activity, they may activate a similar H/K ATPase in the salivary gland or reduce water brash leading to less vagal reflux for saliva production. This was supported by Fornari et al ²⁶ who found elderly people using continuous medication for the gastrointestinal tract were more likely to have xerostomia.

Vitamin D was the third most prevalent medication among patients with xerostomia. Evidence regarding a potential xerostomic mechanism are limited. An in vivo study of rats found that vitamin D may be necessary, along with extracellular calcium, for the secretion of saliva. ²⁷ There is need for further investigation of vitamin D as a causative agent of xerostomia.

The 0.98% prevalence of xerostomia among patients in this academic dental clinic is lower than the global prevalence of 22%. ¹¹ This may be due to inherent limitations in our approach to gathering xerostomia status. Although the method to identify patients with xerostomia was robust, some patients may still have been missed. In addition, the clinic does not assess xerostomia status with a validated xerostomia tool at every visit. Our definition of a xerostomic patient is different than that used by the meta-analysis of Agostini et al. ¹¹ Most studies in the meta-analysis included patients who were greater than 65 years of age, while our study permitted a younger demographic, which could be more reflective of global xerostomia.^11,28

A greater percentage of patients experiencing xerostomia were female and above the age of 65. When encountering patients with these characteristics in practice, pharmacists may consider inquiring about xerostomia status when counseling or performing a medication review. If the patient is experiencing xerostomia, this may be an opportunity to open a dialogue with the prescriber.

When communicating with prescribers regarding interventions to address xerostomia, pharmacists can emphasize the deleterious effects dry mouth can have on oral health and quality of life. Pharmacists may also use evidence-based treatment guidelines and resources regarding potentially inappropriate medications to strengthen an argument to discontinue or switch a xerostomic medication. For a patient receiving a proton-pump inhibitor, referencing the patient’s xerostomia and the Beers Criteria for potentially inappropriate medication in the elderly could start the conversation to discontinue use. ²⁹ The Beers Criteria could also be employed when discussing the necessity of continued use of opioids or antidepressants. For patients with mental health conditions that may be very sensitive to any changes of their prescription regimen, nonpharmacologic (artificial saliva) or pharmacologic measures (cevemiline, pilocarpine) can be suggested to manage xerostomia. Regardless of outcome, xerostomia provides an opportunity for pharmacists to make interventions on behalf of their patients.

The study was limited by the lack of a control group and additional medical history. A control group without xerostomia and matched medical history would have been valuable in determining a correlation between medication classes and xerostomia incidences. Sjogren’s and history of radiation were accounted for as confounders, but a number of other conditions such as rheumatoid arthritis or tobacco use could also contribute as a nonmedication cause of xerostomia. The medication data may potentially contain innaccuracies if the patient is unable to correctly verify what medications they are currently taking. Medication data may also be incomplete if the patient does not have prescription claims data in the external database. For this first attempt using this data source, a modest set of research goals were predetermined. As a result, this study only included data from a single academic clinic. This may limit applicability of these findings to other dental practice settings or clinics. Future studies will be designed with a larger sample size and include a control group to better account for other patient factors that could contribute to xerostomia.

Conclusion

This study examined the medication regimens of patients with xerostomia of an academic dental clinic. The most prevalent xerostomic medication classes were antidepressants, gastric medications, vitamin D, beta-blockers, and opioids. Despite observed prevalence of xerostomia lower than global prevalence, xerostomic medication burden for patients experiencing xerostomia was high. Additional studies are needed to better quantify the degree of xerostomic medication exposure associated with increased risk of experiencing xerostomia. Pharmacist-led interprofessional collaborations should be investigated as a potential avenue to reduce patient xerostomic burden.

Appendix

Veterans Affairs Drug Classification System for Select Drug Classes.

CN609, Antidepressant	MAPROTILINE, TRAZODONE, FLUOXETINE, BUPROPION, SERTRALINE, PAROXETINE, VENLAFAXINE, NEFAZODONE, FLUVOXAMINE, MIRTAZAPINE, CITALOPRAM, ESCITALOPRAM, MILNACIPRAN, VILAZODONE, DULOXETINE, VORTIOXETINE, LEVOMILNACIPRAN, DESVENLAFAXINE, ESKETAMINE
GA900, Gastric Medications	SULFASALAZINE, URSODIOL, SIMETHICONE, OLSALAZINE, ASAFETIDA/MAGNESIA, ORLISTAT, RABEPRAZOLE, ALOSETRON, PANTOPRAZOLE, BALSALAZIDE, OMEPRAZOLE, LANSOPRAZOLE, LUBIPROSTONE, ESOMEPRAZOLE, OMEPRAZOLE/SODIUM BICARBONATE, METHYLNALTREXONE, ALVIMOPAN, ALPHA-GALACTOSIDASE, DEXLANSOPRAZOLE, CARGLUMIC ACID, CHENODEOXYCHOLIC ACID, TEGASEROD, LINACLOTIDE, TEDUGLUTIDE, GLYCEROL, PHENYLBUTYRATE, VEDOLIZUMAB, NALOXEGOL, CHOLIC ACID, DEXTRANOMER/HYALURONATE, ELUXADOLINE, OBETICHOLIC ACID, PLECANATIDE, NALDEMEDINE, PRUCALOPRIDE
VT509, Vitamin D	PARICALCITOL, DOXERCALCIFEROL, VITAMIN D, CHOLECALCIFEROL
CV100, Beta Blockers	NADOLOL, PROPRANOLOL, TIMOLOL, ACEBUTOLOL, METOPROLOL, ATENOLOL, PINDOLOL, LABETALOL, ESMOLOL, PENBUTOLOL, SOTALOL, BISOPROLOL, CARVEDILOL, NEBIVOLOL
CN101, Opioid Analgesics	CODEINE, METHADONE, MORPHINE, LEVORPHANOL, ACETAMINOPHEN/CODEINE, ASPIRIN/OXYCODONE, ACETAMINOPHEN/OXYCODONE, MEPERIDINE, HYDROMORPHONE, BELLADONNA/OPIUM, BUTORPHANOL, PENTAZOCINE, NALOXONE/PENTAZOCINE, ASPIRIN/CARISOPRODOL/CODEINE, ACETAMINOPHEN/HYDROCODONE, OXYCODONE, FENTANYL, ASPIRIN/BUTALBITAL/CAFFEINE/CODEINE, BUPRENORPHINE, NALBUPHINE, ACETAMINOPHEN/BUTALBITAL/CAFFEINE/CODEINE, SUFENTANIL, ALFENTANIL, TRAMADOL, REMIFENTANIL, HYDROCODONE/IBUPROFEN, BUPRENORPHINE/NALOXONE, BUPIVACAINE/FENTANYL, OXYMORPHONE, BUPIVACAINE/HYDROMORPHONE, TAPENTADOL, MORPHINE/NALTREXONE, OXYCODONE/IBUPROFEN, HYDROCODONE, TRAMADOL/ACETAMINOPHEN, ACETAMINOPHEN/CAFFEINE/DIHYDROCODEINE, HYDROMORPHONE/ROPIVACAINE, ACETAMINOPHEN/BENZHYDROCODONE, OLICERIDINE
CN601, Tricyclic Antidepressant	NORTRIPTYLINE, AMITRIPTYLINE, DOXEPIN, IMIPRAMINE, AMOXAPINE, PROTRIPTYLINE, TRIMIPRAMINE, DESIPRAMINE, CLOMIPRAMINE
AH100-AH109, Antihistamine	PROMETHAZINE, CHLORPHENIRAMINE, DIPHENHYDRAMINE, HYDROXYZINE, CYPROHEPTADINE, DIMENHYDRINATE, BROMPHENIRAMINE, CARBINOXAMINE, CLEMASTINE, DEXCHLORPHENIRAMINE, DOXYLAMINE, LORATADINE, CETIRIZINE, FEXOFENADINE, DESLORATADINE, LEVOCETIRIZINE, CHLORCYCLIZINE, TRIPROLIDINE
GU201, Antispasmodics	FLAVOXATE, OXYBUTYNIN CHLORIDE, TOLTERODINE, TROSPIUM, SOLIFENACIN, DARIFENACIN, FESOTERODINE, MIRABEGRON, VIBEGRON