Tolerability and Healthcare Utilization in Maintenance Hemodialysis Patients Undergoing Treatment for Tuberculosis-Related Conditions

Abstract

Background

The incidence of tuberculosis in end-stage renal disease is significantly higher than the general population. Among those with kidney dysfunction, anti-tuberculosis treatment is associated with increased side effects, but the effect on healthcare utilization is unknown.

Methods/Aim

To assess patient-reported symptoms, adverse effects and describe changes in healthcare utilization patterns during treatment for tuberculosis, we conducted a case series (n=12) of patients receiving maintenance hemodialysis from Mayo Clinic Dialysis Services and concurrent drug therapy for tuberculosis from January 2002 through May 2014. Healthcare utilization (hospitalizations and emergency department visits independent of hospital admission) was compared before and during treatment.

Results

Patients were treated for latent (n=7) or active (n=5) tuberculosis. The majority of patients with latent disease were treated with isoniazid (n=5, 71%), while active-disease patients received a 4-drug regimen. Adverse effects were reported in 83% of patients. Compared to measurements prior to drug initiation, serum albumin and dialysis weights were similar at 3 months. Commonly reported anti-tuberculosis drug toxicities were described. More than half (58%) of patients were hospitalized at least once. No emergency department or hospital admissions occurred in the period prior to drug therapy, but healthcare utilization increased during treatment in the latent disease group (hospitalization rate per person-month: pre, 0 vs. post, 1).

Conclusions

Among hemodialysis patients, anti-tuberculosis therapy is associated with frequent patient-reported symptoms and increased healthcare utilization. Patients receiving treatment for latent disease may have the greatest increase in healthcare use. Careful monitoring and early complication detection may help optimize medication adherence and minimize hospitalizations.

INTRODUCTION

According to the 2014 World Health Organization (WHO) Global Report, tuberculosis (TB) remains a major public health concern. It is estimated that worldwide, 9 million people developed TB in 2013 and 1.5 million died from the disease [1]. A particularly susceptible group of patients are those with end-stage renal disease (ESRD) receiving maintenance hemodialysis (HD) therapy[2]. The risk for dialysis patients developing TB can be up to 52 times higher than the general population depending on the geographic region [3-5]; the prevalence of TB may range from 1.6% to 5.8% in developed countries and upwards of 28% in some developing countries [4]. The increased disease acquisition in ESRD is likely multifactorial but often attributed to decreased cellular immunity in the uremic state and a higher prevalence of high-risk comorbidities such as advanced age, diabetes, and low body mass index [4,6-10]. Importantly, survival is limited following TB, with mortality rates ranging from 17% to 75% among dialysis-dependent patients [4,11]. In this patient group, clinical presentations of pulmonary TB disease may be atypical, given the high prevalence of extrapulmonary disease [12-14]. In addition, non-specific symptoms which may be masked by adverse effects from uremia and/or chronic dialysis therapy may also conceal underlying TB disease [3,10,15]. Therefore, high mortality rates are often ascribed to delayed disease recognition leading to late diagnosis and subsequent treatment [3,16]. Additionally, ESRD increases risk of latent TB progression to active TB [17,18].

Given the increased risk of TB acquisition among ESRD patients and prolonged time spent in healthcare settings which may expose other high-risk persons, the Centers for Disease Control and Prevention (CDC) has recommended that patients requiring maintenance dialysis should have at least one test for TB infection, with subsequent screening in the event of exposure, to aid in the detection of latent tuberculosis infection (LTBI) or subclinical disease [19 ]. However, there is no uniform screening approach for patients on dialysis [20]. Common amongst HD patients (20% to 70%), LTBI is defined as a reactive tuberculin skin test (TST) or positive interferon gamma release assay (IGRA) in the absence of any clinical or radiographic evidence of active TB disease [16]. Treatment of LTBI requires isoniazid, rifampin, or combination isoniazid plus rifapentine, administered over 3-9 months depending upon the specific drug(s) utilized [21]. Previously, 8 weeks of combination rifampin plus pyrazinamide was recommended as an alternative LTBI treatment option; however, due to higher rates of drug-associated hepatoxicity, this combination regimen was no longer recommended after 2003 [22]. Unfortunately, anti-TB medications can produce many toxicities. This is particularly true in those with kidney failure, wherein the clearance of both parent drugs and their metabolites may be altered [12,23,24]. To address this, the American Thoracic Society published recommendations for various medication dosing adjustments in patients with renal dysfunction and those requiring dialysis [25]. Despite prevalent knowledge of the side effect profile of anti-TB treatment, treatment-related healthcare resource utilization such as emergency department (ED) visits and hospitalizations in this population has not been well-described. In this case-series, we examined the major symptoms, adverse effect profile, and frequency and healthcare resource utilization patterns in maintenance HD patients receiving anti-TB therapy at our outpatient HD facilities.

SUBJECTS AND METHODS

Study Overview

The study cohort was identified from the Mayo Clinic institutional database. An initial query for patients (n=2,952) with a diagnosis of any TB or mycobacterial infection was through international classification of disease (ICD) 9 codes (starting with 010, 011, 012, 013, 014, 015, 016, 017, 018, 031, 137) from January 1, 2002 through May 31, 2014. Patients were then cross-referenced with ICD9 and billing codes for end-stage renal disease 585.6, dialysis V56.8, renal dialysis status or dependence on renal dialysis V45.11, or chronic kidney disease codes (n=159). The final cohort (n=12) was restricted to patients receiving maintenance HD therapy within one of Mayo Clinic Dialysis Services’ 8 dialysis units from January 1, 2002 to May 31, 2014 as previously described [26,27]. Electronic chart reviews were conducted to confirm study eligibility and collect study data. Minnesota research authorization was available for all study participants. The Mayo Clinic Institutional Review Board approved the study.

Patient Demographics and Covariables

Patients’ electronic medical records were reviewed for baseline demographics, kidney failure cause, laboratory values, anti-TB medications, changes to anti-TB therapy, documented adverse effects or new symptoms while on TB treatment, and ED visits and hospitalizations.

Definitions

Active TB infection was defined as a positive culture of any tissue or specimen for Mycobacterium tuberculosis; or a positive acid-fast bacilli (AFB) stain in the presence of typical symptoms and absence of atypical mycobacterial infection. A diagnosis of LTBI was based on a positive standard purified protein derivative (PPD) skin test or interferon gamma release (QuantiFERON; QFT®) test and no concomitant signs or symptoms of active infection or radiographic evidence of pulmonary TB. At our institution, the policy is to screen hemodialysis patients for LTBI at the initiation of dialysis with either PPD or QuantiFERON testing. Otherwise, only patients on the kidney transplant waiting list were screened annually during their transplant candidate evaluations. Active TB infection and LTBI treatment regimen selection did follow a standard protocol as outlined by Center for Disease Control [22,25,28]. Treatment for latent TB consisted of 6 or 9 months of single agent isoniazid, rifampin for 4 months, or 8 weeks of combination rifampin plus pyrazinamide which was an acceptable regimen prior to modified recommendations in 2003. Treatment of active TB was with four-drug regimen of isoniazid, rifampin, ethambutol, and pyrazinamide (continuation phase) then isoniazid and rifampin for 2 months (continuation phase) for a total of 6 months. If necessary, the above regimens were modified at the discretion of the treating infectious disease specialist if toxicity or medical complications arose.

Symptoms experienced during anti-TB therapy were obtained through chart review of outpatient dialysis, specialty, or primary care provider visits, dialysis nursing notes from each treatment, and hospital or ED visits. Signs and symptoms assessed by providers or reported by patients, particularly those leading to an ED visit or hospital admission, were recorded during the course of anti-TB therapy. Signs and symptoms which were present prior to anti-TB therapy were not recorded unless the intensity of the prior symptom was magnified during treatment course. Common toxicities of anti-TB therapy were recorded (when documented) including hepatitis (liver enzyme elevation), nausea, vomiting, peripheral neuropathy, peripheral blood cytopenias, skin rash, and arthralgias. Other complaints were also recorded, such as lightheadedness or dizziness and syncope. Elevated liver enzymes were defined as values of alanine aminotransferase (ALT) or aspartate aminotransferase (AST) above the upper limit of normal.

The primary outcome was healthcare utilization. Healthcare utilization was broadly defined as need for hospital admission or ED care without hospitalization during the period of anti-TB therapy. Evaluations in the ED leading to hospitalization were counted as a hospital admission only. ED evaluations which resulted in dismissal from the ED were categorized as such. Hospitalization and ED utilization rates were calculated for each patient. To compare healthcare utilization patterns before and during treatment, hospitalization and ED visits were determined in the 6 months prior to treatment and during treatment. Overall, patient follow-up was restricted to the time the patient was both receiving anti-TB therapy and undergoing maintenance HD therapy at Mayo Clinic Dialysis Services (MCDS). Therefore, patient follow-up ended at completion of anti-TB therapy, time of transfer to non-MCDS dialysis unit, or discontinuation of dialysis (e.g., kidney transplantation).

Statistical Analysis

Data were expressed as mean and standard deviation for normally distributed data. Data that were not normally distributed were expressed as median and interquartile range. Event rates were defined as number of events per period at risk (per-person-months). Statistical tests were not performed due to the small number of patients and frequency of multiple events in select patients.

RESULTS

Overall Cohort

From January 2002 through May 2014, 12 patients receiving maintenance, in-center HD in Mayo Clinic Dialysis Services were treated for LTBI (n=7) or active TB (n=5). Patient characteristics and drug therapy are provided in Table 1. The median age at treatment was 60 years (range 24-86), 75% were male, 50% were white, mean dialysis vintage was 15.3 months, diabetes was the cause of ESRD in 50%, mean weight at dialysis was 83 kg (range 45-110), and mean serum albumin was 3.2 g/dL (range 2.1-4.2). In the group of patients who were diagnosed with active TB, 60% had pulmonary TB and 40% had extra-pulmonary TB. The majority (71%) of LTBI patients were treated with isoniazid monotherapy. As per the published CDC guidelines, 100% of HD patients taking isoniazid also received pyridoxine [28]. Major symptoms, adverse effects, and healthcare utilization patterns during anti-TB therapy are shown in Table 2. All Active TB patients reported at least one adverse effect, whereas 5 of 7 (71 %) LTBI patients experienced adverse effects. The most frequent adverse effects overall were syncope (42%), lightheadedness/dizziness (42%), peripheral neuropathy (25%), and liver enzyme elevation or abnormal liver function tests (LFTs: 17%). These adverse effects led to therapy modifications in only 3 patients (25%). Over the initial course of treatment, serum albumin and weight changes were minimal with the exception of patient #7, who lost 8 kg (Table 3). Over half (58%) of patients were hospitalized at least once (median visits 1; range 1-3). However, ED evaluations without a hospital stay for complications or symptoms experienced during the treatment period were common (25%; median visits 1; range 1-3), particularly in the LTBI patient group. The overall number of hospitalizations, ED visits (only), and anti-TB medication changes during treatment are illustrated in Figure 1 for LTBI and active TB patient groups.

Table 1.

Patient characteristics and anti-TB drug therapy in chronic hemodialysis patients with Latent and Active tuberculosis.

Patient	Age (years)	sex	Dialysis Vintage (months)	ESRD cause	Treatment duration (months)	Isoniazid*	Pyrazinamide*	Rifampin*	Ethambutol*	Treatment change or discontinuation of original prescription
Latent TB
1	52	M	26	DM	9	300 D	-	-	-	No
2	59	M	1	DM	1.5¥	300 D	-	-	-	Yes
3	63	M	45	Unknown	4	300 D	-	-	-	No
4	71	F	47	DM	9	300 D	-	-	-	No
5	86	M	7	HTN	9	300 D	-	-	-	No
6	57	M	0¶	HTN	2	-	2000 3W	600 D	-	No
7	48	M	10	DM	4	-	-	600 D	-	No
Active TB
8	24	M	3	Unknown	6	900 3W	1500 2W	600 2W	1200 3W	No
9	61	M	−3~	ATN	6	600 D	2000 D	900 D	1600 D	No
10	66	F	8	DM	1.5¥	900 3W	2500 3W	600 3W	2000 3W	Yes
11	84	M	7	DM	6	300 D	1500 3W	600 D	1000 3W	No
12	43	F	30	GN	6	300 D	1000 3W	600 D		Yes

Abbreviations: DM, diabetes mellitus; ESRD, end-stage renal disease; F, female; GN, glomerulonephritis; HTN, hypertension; M, male.

Drug dosage is reported in milligrams.

^*If dose information not reported, the patient did not receive this medication as a part of the regimen

^‡ Freq: Frequency (D: daily, 2W: twice weekly, 3W: thrice weekly),

Standard treatment regimen options for LTBI during the study period: Isoniazid for 6 or 9 months, Rifampin for 4 months, Rifampin and Pyrazinamide (combination no longer supported by the Center for Disease Control beginning in 2003 due to reports of liver injury) for 2 months duration. http://www.cdc.gov/tb/topic/treatment/ltbi.htm

Standard treatment regimen options for Active TB during the study period included combination regimens of Isoniazid, Rifampin, Pyrazinamide, and Ethambutol (initial phase) followed by Isoniazid and Rifampin (continuation phase). http://www.cdc.gov/tb/topic/treatment/tbdisease.htm

Dose adjustments and other alterations: Pyrazinamide and Ethambutol require dosing adjustments for renal insufficiency. Isoniazid is provided at standard dosing regimens of either 300 mg/day or 900 mg thrice weekly.,

Patient specific information:

^¶dialysis started at same week of diagnosis of TB,

^~TB was diagnosed 3 months before starting dialysis but patient continued to receive treatment while on dialysis.

^¥The duration reflects the time patients were undergoing both chronic hemodialysis and treatment for TB. A shorter treatment duration indicates that patients received transplant or were no longer on dialysis.

Table 2.

Symptomatology, adverse effects and healthcare utilization during anti-tuberculosis therapy.

Patient	Anti-TB drugs	Major Symptoms or Adverse Effects	Other Symptoms or Adverse effects	Hospitalization (N)	Cause of hospitalization	ED Visits (N)	Cause of ED visits
Latent TB
1	INH	Seizure	Lightheadedness	Yes (3)	Subdural hematoma, seizures, hyperkalemia	No	-
2	INH	Syncope	Lightheadedness, Elevated LFTs	Yes (2)	Abdominal Pain	Yes (3)	Abdominal pain
3	INH	Peripheral neuropathy	-	No	-	No	-
4	INH	Syncope	Muscle weakness Dizziness, Elevated LFTs, N/V	No	-	No	-
5	INH	-	-	No	-	No	-
6	PZA, Rif	Syncope	Rash, lightheadedness	Yes (1)	Syncope	Yes (1)	Syncope
7	Rif	-	-	No	-	No	-
Active TB
8	INH, Rif, PZA, Eth	Chest pain	N/V	Yes (1)	Chest pain	No	-
9	INH, Rif, PZA, Eth	Syncope	-	Yes (1)	Syncope	No	-
10	INH, Rif, PZA, Eth	Peripheral Neuropathy	Muscle weakness Confusion	No	-	No	-
11	INH, Rif, PZA, Eth	Syncope	-	Yes (1)	Syncope	Yes (1)	Syncope
12	INH, Rif, PZA	Peripheral Neuropathy	Lightheadedness, muscle weakness, Elevated LFTs, N/V	Yes (3)	Muscle Weakness, Pancreatitis	No	-

Abbreviations: TB, tuberculosis, Eth, Ethambutol; INH, Isoniazid; LFTs, liver function tests; N/V,n, nausea and vomiting; PZA, Pyrazinamide; Rif, Rifampin ED visits are emergency department visits that did not lead to admission, which is otherwise reported under hospitalization. N: number of hospitalizations or ED visits.

Table 3.

Healthcare utilization patterns before and after anti-tuberculosis therapy initiation.

Patient	Age (years)	sex	Dialysis Vintage (mos)	Therapy duration (mos)	Weight change at 3 mos (kg)	Serum albumin change at 3 mos (g/dL)	Hospitalization pre- therapy* (N)	Hospitalization during TB therapy *(N)	#Hospital days during therapy	Hospitalization RATE during therapy	ED pre- therapy* (N)	ED visits during therapy* (N)	ED RATE during therapy
Latent TB
1	52	M	26	9	0.7	−1.3	No	Yes (3)	64	0.014	No	No	0
2	59	M	1	1.5	−0.6	1.2	No	Yes (2)	11	0.059	No	Yes (3)	0.07
3	63	M	45	4	1.9	n/a	No	No	0	0	No	No	0
4	71	F	47	9	1.0	−0.1	No	No	0	0	No	No	0
5	86	M	7	9	3.5	−0.2	No	No	0	0	No	No	0
6	57	M	0¶	2	2.2	n/a	Unknown	Yes (1)	4	0.018	Unknown	Yes (1)	0.017
7	48	M	10	4	−8.0	0.5	No	No	0	0	No	No	0
Total Latent TB	-	-	-	38.5	0.1	0.02	0	6	79	0.0056	0	4	0.0034
Active TB
8	24	M	3	6	2.5	0.6	Yes (4)	Yes (1)	2	0.0056	No	No	0
9	61	M	−3~	6	10.6	n/a	Yes (1)	Yes (1)	69	0.0090	No	No	0
10	66	F	8	1.5	0.5	n/a	No	No	0	0	No	No	0
11	84	M	7	6	0.7	−0.6	No	Yes (1)	7	0.0058	Yes (1)	Yes (1)	0.0056
12	43	F	30	6	0	0.7	Yes (1)	Yes (3)	13	0.018	No	No	0
Total Active TB	-	-	-	25.5	2.86	0.002	6	6	91	0.0089	1	1	0.0013
Entire cohort	-	-	-	64.0	1.25	0.1	6	12	170	0.0069	1	5	0.0026

TB: tuberculosis. ED: visits are emergency department visits that did not lead to admission, which is otherwise reported under hospitalization. LFTs: Liver function tests. Mos: months. N/V: Nausea and vomiting, N: number of hospitalizations or ED visits. Rate is expressed as ED visits (hospitalization) -per-person-day. (To change to per-person-month multiply these results *30)

^*In the 6 months prior to initiation of anti-TB treatment, for those who have been on dialysis less than 6 months, only admissions/ED visits during the period on dialysis was included.

^~Patient was not followed at our institution

^¶dialysis started at same week of diagnosis of TB

Figure 1.

Number of hospitalizations, emergency department visits (only), and anti-tuberculosis medication change among dialysis patients receiving treatment for Latent (n=7) or Active (n=5) tuberculosis

Major Symptoms and Adverse Effects Profile

LTBI group

Among those treated for LTBI (n=7), anti-TB treatment consisted of isoniazid monotherapy (n=5) or was rifampin based (n=2). During anti-TB therapy, the most frequent complaints were lightheadedness (43%) and syncope (43%), which often lead to additional healthcare use. One LTBI patient (Patient #2) discontinued treatment after 6 weeks due to elevated LFTs attributed to isoniazid toxicity. Only one LTBI patient reported no adverse symptoms and appeared to tolerate the treatment regimen.

Active TB group

Among those treated for active TB (n=5), anti-TB therapy was comprised of a four-drug isoniazid-based regimen. The most frequent patient-reported symptoms during anti-TB therapy were peripheral neuropathy (40%), syncope (40%), and light headedness (20%). One patient in this group developed confusion. Two Active TB patients (40%) required a medication change or discontinuation of treatment. Patient #10 developed severe peripheral neuropathy and muscle weakness 6 weeks after starting treatment with rifampicin, isoniazid, ethambutol, and pyrazinamide and opted to temporarily discontinue the treatment. Patient #12 developed significant peripheral neuropathy and elevated LFTs. His isoniazid and ethambutol were held more than once during the treatment course, and then reintroduced at a later date, with a 6-month total duration of therapy.

Healthcare Utilization Before and After Anti-TB Therapy

Healthcare utilization defined as hospitalization or ED visit without hospital admission are illustrated in Figures 2a and 2b by TB treatment group and timing of TB therapy. The LTBI patients had no ED visits or hospitalizations prior to anti-TB therapy, but these events increased after treatment was started. While most patients had new onset of symptoms during anti-TB therapy, 3 patients (2-LTBI and 1-Active) had the highest healthcare utilization rates and accounted for the majority of hospital and ED visits, as illustrated in Table 3.

Figure 2a.

Number of hospitalizations 6 months prior to initiation and during anti-tuberculosis therapy by Active (n=5) and Latent (n=7) TB groupings

Figure 2b.

Number of emergency department visits (without hospitalization) 6 months prior to initiation and during anti-tuberculosis therapy by Active (n=5) and Latent (n=7) TB groupings

LTBI group

Among the LTBI patients, 43% were hospitalized (median visits 2; range 1-3) and 26% were evaluated in the ED (median visits 2; range 1-3) for complications or symptoms experienced during the treatment period. In the preceding 6 months, LTBI patients had no hospitalizations or ED visits. However, after initiation of LTBI therapy, these events increased. Notably, Patient #2 received isoniazid monotherapy and was hospitalized twice and had 3 additional ED visits. His complaints were primarily gastroenterological, such as abdominal pain and constipation, but 2 hospitalizations were for syncope, lightheadedness, and elevated LFTs. His treatment was aborted before completion. Despite the high healthcare utilization rate, only one patient had a medication change in the LTBI group. Another patient (Patient #1) at 14 weeks into treatment was hospitalized a total of 64 days after developing a subdural hematoma complicated by recurrent seizures and physical debility. This patient accounted for 50% of the total LTBI group hospitalizations.

Active TB group

Among the Active TB patients, a majority (80%) was admitted to the hospital at least once (median visits 1; range 1-3). Presenting symptoms and diagnoses upon hospital admissions included syncope, muscle weakness, pancreatitis, and chest pain. In the Active TB group, total numbers of hospital and ED visits were similar in the 6 months prior to and during treatment for TB. However, Patient #12 had the highest hospitalization rate (0.02 per-patient-day or 0.54 per-person-month) and was admitted to the hospital 3 times during the treatment period, accounting for 50% of the total Active TB group hospitalizations. Muscle weaknesses, peripheral neuropathy, abdominal pain, light headedness, and elevated LFTs were all reported. The patient was eventually diagnosed with pancreatitis later complicated by recurrent pancreatitis episodes with no clear etiology.

DISCUSSION

Our study reveals for the first time a pattern of increased healthcare utilization relative to pretreatment period and poor anti-TB therapy tolerance among patients receiving maintenance HD. In particular, patients treated for LTBI had the most significant change in number of hospitalizations and ED visits when compared to the 6-month period directly preceding anti-TB treatment. Overall, only two of 12 (17%), were free of patient-reported symptomatology during therapy. New onset of syncope, lightheadedness, and neuropsychiatric toxicities accounted for 50% of complaints leading to hospitalization, and abdominal pain accounted for 60% of ED visits during the treatment period. However, other serious events, including pancreatitis and subdural hematoma, occurred. These findings illustrate the potential impact of anti-TB treatment on acute care service needs due to the toxicities of anti-TB therapy and/or heightened sensitivity to symptoms typically tolerated during the course of maintenance HD therapy.

The captured major symptom and adverse effect profile of patients in our study is similar to previous reports. Quantril et al. [12] reported a high incidence of adverse effects related to TB medications leading to changes in drug treatment in 16 dialysis patients (14 peritoneal dialysis; 2 HD) treated for Active TB. In comparison, in our study only 25% of subjects were clinically deemed to require a medication change or discontinuation, although those with frequent acute care services use may have benefited from changes in therapy. Quantril and group also reported neuropsychiatric side effects in 2 patients in the form of confusion and seizure activity potentially related to isoniazid neurotoxicity [12,23]. In our study, syncope, lightheadedness, and neuropsychiatric effects accounted for 50% of hospitalizations, and abdominal pain accounted for 60% of ED visits. Syncope and lightheadedness were common among our cohort, which may relate to decreased tolerability of the hemodynamic changes and volume removal during HD despite minimal changes found in body weight during the treatment period. Notably, one patient developed acute pancreatitis, a complication that has been described with isoniazid [29]. From a healthcare utilization perspective, over one-half (58 %) of our patients had at least one hospitalization. In comparison, in the 6 months prior to treatment, hospitalizations were non-existent for the LTBI group, illustrating the increased usage of acute services. However, 3 patients accounted for 8 hospitalizations and 3 ED visits, illustrating that some patients were less tolerant of adverse effects, leading to an increased need for acute care services. Overall, these findings suggest that in maintenance HD patients drug-related toxicities are common and of significant impact.

Given the increased risk of side effects in the dialysis population, the American Thoracic Society, the Infectious Diseases Society of America, and the CDC have jointly published guidelines for the appropriate dosing and dosing intervals in this population to improve tolerance and decrease incidence of side effects [25].. As a group, patients with LTBI appeared to experience fewer symptoms and adverse effects during the study period, likely due to monotherapy with isoniazid or rifampin. However, some LTBI patients appeared to have poor drug tolerance leading to increased healthcare utilization, suggesting an exaggerated sensitivity to toxicities of anti-TB therapy beyond that of other HD patients.

Our case series has some limitations. The small number of patients has limited the ability to fully explore the spectrum of drug toxicities and healthcare utilization patterns associated with treatment in the HD population. Patients receiving maintenance HD represent a complex, high-risk population for morbidity and mortality. Therefore, symptom assessment may be blurred by the multiple comorbidities and natural course of underlying chronic disease progression such as diabetes. Future studies with a control group of dialysis and non-dialysis patients may allow for comparison of the true magnitude of adverse drug reactions and healthcare utilization differences attributed to advanced chronic kidney disease and dialysis-dependence. Understanding this limitation, we compared hospital and ED visit numbers pre-anti-TB therapy to the number during treatment to highlight differences related to initiation of therapy. While larger studies of HD patient populations would be useful to calculate healthcare utilization rates, if these same studies apply billing claims data, they may help confirm trends but will lack the granularity of detail which can be obtained by individual chart review.

Overall, our findings raise concerns not only regarding the negative impact on the health of the patients, but also their quality of life, particularly during times requiring increased healthcare use and the increased cost of care attributed to anti-TB treatment. In particular, this may be of concern to outpatient dialysis facilities with a higher incidence of LTBI. Given the recent implementation of the Center for Medicare and Medicaid Services (CMS) Quality Incentive Program, hospital readmissions are being included as quality measures in outpatient dialysis facilities [30,31]. Based on the increased frequency of healthcare utilization found in our study, facilities with high LTBI or Active TB rates may have a disadvantage due the potential risk for increased hospitalization (and 30-day readmission) rates in their facilities. LTBI or Active TB could be an important patient comorbidity risk adjustor when comparing the hospitalization and readmission rates. In addition, those requiring anti-TB treatments may be denied admission to dialysis facilities because of their increased resource utilization. However, further studies are needed to assess the impact of anti-TB therapy on morbidity, mortality, and healthcare utilization in this population.

In conclusion, anti-TB therapy is associated with frequent symptoms and adverse effects leading to increased healthcare resource utilization in maintenance HD patients. Increased awareness of potential complications and careful monitoring of patients is essential to optimize medication adherence, recognize symptoms of intolerance early, and understand when adjustments may be needed to dialysis treatments or medication regimens given symptomatology. Collectively, these efforts may help minimize healthcare utilization for acute care needs due to adverse effects of anti-TB therapies.