Outcomes and Safety of Outpatient Parenteral Antimicrobial Therapy in Select Children with Cystic Fibrosis

Abstract

Background: Pulmonary exacerbations (PExs) are common in individuals with cystic fibrosis (CF). Data regarding outcomes of outpatient parenteral antimicrobial therapy (OPAT) in children are sparse.

Methods: Retrospective data of PEx episodes treated in the hospital versus OPAT collected. Children ≤18 years were included. Outcome measures included FEV₁, FVC, FEF_25–75%P, time to the next PEx, and weight gain.

Results: Eighty-three subjects with 290 PEx events were eligible. The hospital group had 242 and the OPAT group had 48 PEx events. The median age was 13.1 years for the OPAT and 13.4 years for the hospital group. Medicaid coverage was higher in the hospital group (82.2%) versus OPAT group (48.9%, P < 0.0001). The hospital group had lower FEV₁%P on admission [72%P (interquartile range [IQR] = 59.7 and 84) versus 80%P (IQR = 70.7 and 89); P = 0.001] and at the end of treatment [86%P (IQR = 72 and 96.7) versus 92%P (IQR = 82 and 101); P = 0.003] in comparison with OPAT group. FEV₁%P improved more in the hospital group, [12%P (IQR = 4 and 20)] versus in the OPAT group [8%P (IQR = 2 and 22.5); (P = 0.41)] but did not quite reach a statistically significant level. The hospital intravenous (IV) group gained more weight (P < 0.0001). There was no difference between the 2 groups in time to the first PEx (P = 0.47) and adverse events.

Conclusion: OPAT was safe and comparable with hospital therapy in a select group of children with CF. Hospital IV should be considered for sicker children and families with limited resources.

Introduction

Chronic airway inflammation and infection leading to progressive lung disease is the major cause of morbidity and mortality in individuals with cystic fibrosis (CF). Pulmonary exacerbations (PExs) are common in CF, and it has been shown that in 25% of CF patients, pulmonary functions fail to recover to the baseline level after treatment.^1,2 Because each PEx has a potential negative impact on pulmonary health, effective treatment of exacerbations is essential to maintain lung health in CF. Antimicrobial treatment has been the basis for treatment of CF PEx.³ Although oral and inhaled antimicrobials provide adequate treatment for mild to moderate PEx in the outpatient setting, intravenous (IV) antimicrobials are often required for the treatment of severe PEx.^1–3

Outpatient parenteral antimicrobial therapy (OPAT) is defined as the administration of parenteral antimicrobial therapy with a minimum of 2 doses on separate days and does not require hospitalization.⁴ Since the initial use of OPAT as an alternative to hospital therapy in the 1970s, multiple studies have shown treatment success in non-CF populations.⁵ Although this is true in non-CF populations, studies investigating the effectiveness and clinical outcomes of OPAT in individuals with CF have conflicting results.^6–12

The first prospective controlled study to compare the efficacy and benefits of OPAT versus hospital treatment found that both OPAT and hospital treatments had significant improvement in pulmonary function.¹³ A single prospective randomized study conducted in 1997 with a sample size of 17 subjects did not show a significant difference between OPAT and hospital groups with respect to body weight, 12-min walking distance, sputum, weight, pulse oximetry, and improvement in lung function.¹⁴ Two retrospective pediatric studies comparing OPAT with hospital treatment showed conflicting results. One favored hospital treatment,¹⁵ whereas the other found OPAT was not inferior to hospital IV for the treatment of CF PEx in children.¹⁶ Recently, data from the Epidemiologic Study of CF determined a significant advantage to inpatient versus OPAT for CF PEx in mixed adult and pediatric patients.¹⁷

With more family-centered care by hospitals, OPAT has become more accessible for pediatric patients with CF. However, CF centers utilize OPAT only in select children due to various reasons, and a significant portion of children with CF complete IV treatment in the hospital. The purpose of this study was to review our center's OPAT experience in select children with CF and investigate the outcomes of OPAT versus hospital therapy for the treatment of PEx.

Materials and Methods

The study included a retrospective review of medical records of children with CF followed at Arkansas Children's Hospital Cystic Fibrosis Care Center. The study was approved by the Institutional Review Board at the University of Arkansas for Medical Sciences. A waiver of informed consent was granted for the study.

Study population

Patients identified for the study had a known diagnosis of CF and were admitted to the hospital owing to PEx between 2000 and 2014. Data collection included gender, age on admission, cystic fibrosis transmembrane conductance regulator (CFTR) mutation, and existence of pancreatic insufficiency. Admission weight and body mass index (BMI) (and percentiles) were extracted and compared with measures obtained after treatment of PEx. Pulmonary function test variables including FEV₁%P, FVC %P, and FEF_25–75%P were obtained on admission day and at the end of treatments. Patients with positive respiratory cultures for methicilline susceptible Staphylococcus aureus (MSSA), methicilline resistant Staphylococcus aureus (MRSA), and Pseudomonas aeruginosa were identified. Existence of G-tube, CF-related diabetes (CFRD), CF liver disease, antibiotic-related adverse reactions, and admission and post-treatment biochemistry and liver function tests were recorded.

Inclusion/exclusion criteria

Patients ≤18 years of age who could perform reliable spirometry were included. OPAT eligibility was determined at the CF center's discretion and the major drivers of how and where patients were to be treated were based on the following criteria: (1) Patients' age—young children (<6 years old) were treated in the hospital. (2) Family resources—when families had limited resources and the caregivers were not comfortable with the treatment at home, their children were treated in the hospital. (3) Severity of the symptoms—if children had severe symptoms such as requiring O₂ or ventilatory support, IV treatment completed in the hospital. (4) Complexity of the treatment—if patients required multiple antibiotics and/or nephrotoxic drugs requiring frequent dosing adjustment, they were treated in the hospital. (5) Availability of home health agencies to help with the central line care and supervision of home antibiotics. (6) History of drug allergies requiring desensitization or history of adverse reactions to the drugs. (7) If a patient/family had a history of poor adherence with treatments, the treatment was completed in the hospital. (8) Some families lived in places that were 5–6 h away from the hospital. These children's IV treatment was completed in the hospital. In summary, OPAT was utilized mostly in relatively healthy children (pulmonary, nutrition, mental health, etc.) and in families with adequate resources. OPAT was defined as treatment administered in the hospital for an initial 2–3 days, after which, the full treatment course was completed at home (10–14 days). Hospital IV was referred to a full IV treatment course administered in the hospital. Patients presenting with acute respiratory failure, pneumothorax, hemoptysis, and on the active lung transplant listing were excluded.

Outcome measures

Primary outcome measures included improvement in FEV₁%P (expressed as arithmetical difference in FEV₁%P before and after treatment), FVC %P, and FEF_25–75%P. A successful treatment was defined as improvement in symptoms and FEV₁ to baseline measurement (within 90% of the best FEV₁ during the past 12 months). Time until the next PEx was defined as the time period from the end of the treatment (either at home or in the hospital) to a new PEx. Secondary outcome variables included changes in weight, BMI, and BMI percentiles before and after treatment.

Statistical analysis

As described in a previous publication,¹⁶ our data represented number of PEx episodes and not number of patients. Therefore, there were sick children who required multiple hospitalizations each year that would represent a significant portion of episodes and possibly skewing our data. To prevent overrepresenting some patients with multiple PEx, we included only the first PEx of these children per year to the statistical analysis. In addition, to exclude the variable of experience with PEx treatment, we included PEx of children with at least 4 previous PEx treatments (OPAT, hospital).¹⁶ Descriptive statistics were calculated for study variables including median, first quartile, and third quartile for continuous variables and frequency with percentages for categorical variables. Comparison of distributions for continuous variables between home IV and hospital IV treatment was performed using Wilcoxon rank sum test. Comparison of proportions for categorical variables was performed using chi-square tests or Fisher's exact test for any frequency ≤5. Results are reported as P values and odds ratios (ORs) with 95% confidence intervals (CIs). OR and CI >1 indicated increased odd of OPAT in case versus control group. OR and CI <1 indicated that OPAT was unlikely in the case group. All values of P < 0.05 were considered to indicate statistical significance. The data were analyzed using statistical software R v.3.2.3 (R development Core Team, Vienna, Austria).

Results

Study population

Eighty-three children met the inclusion criteria. The majority of children had >1 PEx (1–15) during the study period. Each PEx was recorded as an independent entry resulting in a total sample size of 290 PEx events. Of the 290 PEx treatment courses, 242 were completed in the hospital (hospital IV group) and 48 were completed at home (OPAT group). The number of children treated with OPAT increased some over years, but it comprised only a small fraction of IV antimicrobial courses (<20%/year). The median age of children was 13.1 years [interquartile range (IQR) = I9.6–15.1 years] for the OPAT group and 13.4 years (IQR = 10.1–15.8 years) for the hospital IV group. There was no statistical difference between hospital and OPAT groups regarding gender, CFTR mutations, known drug allergy, existence of G-tube, CFRD, respiratory cultures for CF-related pathogens (nonmucoid and mucoid P. aeruginosa, MSSA, MRSA) (Table 1). Although 75% of the OPAT group had previous experience with home IV, only 32.8% of the hospital IV group had previous home IV experience [P < 0.0001; OR 2.37 (1.4–3.8)]. Children who were under Medicaid insurance (a social health program for families with limited resources) were less likely be treated in the hospital [P < 0.0001; OR 0.5 (0.3–0.9)].

Table 1.

Characteristics of the Hospital and Outpatient Parenteral Antimicrobial Therapy Groups

Variable	Hospital IV (N = 242)	OPAT (N = 48)	P	OR (95% CI)
	n (%)	n (%)
Known drug allergy			0.73	1.05 (0.6–1.7)
No	85 (35.1)	15 (31.2)
Yes	157 (64.9)	33 (68.8)
Experience with OPAT			<0.0001	2.37 (1.4–3.8)
No	165 (68.2)	12 (25)
Yes	77 (31.8)	36 (75)
Gender			0.37	0.8 (0.4–1.4)
Female	131 (54.1	22 (45.8)
Male	111 (45.9)	26 (54.2)
Genotype			0.22	1.1 (0.6–1.75)
delF508/del F508	174 (72.2)	38 (80.9)
delF508/other	53 (22)	9 (19.1)
Other/other	14 (5.8)	0 (0)
G-tube			0.07	0.65 (0.3–1.2)
No	134 (55.4)	34 (70.8)
Yes	108 (44.6)	14 (29.2)
CF liver disease			0.1	2.1 (0.8–5.1)
No	223 (92.1)	40 (83.3)
Yes	19 (7.9)	8 (16.7)
CF-related diabetes			0.99	0.9 (0.3–2.2)
No	204 (84.3)	41 (85.4)
Yes	38 (15.7)	7 (14.6)
Mucoid pseudomonas			0.82	0.8 (0.3–2.0)
No	198 (82.8)	41 (85.4)
Yes	41 (17.2)	7 (14.6)
Nonmucoid pseudomonas			0.91	1.08 (0.5–2)
No	174 (72.8)	34 (70.8)
Yes	65 (27.2)	14 (29.2)
MRSA			0.73	0.9 (0.5–1.5)
No	100 (41.8)	22 (45.8)
Yes	139 (58.2)	26 (54.2)
MSSA			0.89	0.9 (0.4–1.8)
No	174 (72.8)	36 (75)
Yes	65 (27.2)	12 (25)
Insurance			<0.0001	0.5 (0.3–0.9)
Medicaid	199 (82.2)	23 (48.9)
Non-Medicaid	43 (17.8)	24 (51.1)

CF, cystic fibrosis; CI, confidence interval; IV, intravenous; MRSA, methicilline resistant Staphylococcus aureus; MSSA, methicilline susceptible Staphylococcus aureus; OPAT, outpatient parenteral antimicrobial therapy; OR, odds ratio.

Outcomes

On admission, the median FEV₁%P for the hospital IV group was statistically lower [72%P (IQR = 59.7–84)] from the OPAT group [80%P (IQR = 70.7–89); P = 0.001]. Similarly, at the end of treatment, the FEV₁%P measurements were statistically lower in the hospital IV group [86%P (IQR = 72–96.7)] versus OPAT group [92%P (IQR = 82–101); P = 0.003]. FEV₁%P values improved more in the hospital group; however, this was not statistically significant [12% (IQR = 4–20) versus 8% (IQR = 2–22.5); P = 0.41]. Children in the OPAT group had a higher median FVC%P values before treatment [92%P (IQR = 82.5–99.5) versus 86%P (IQR = 71.7–97); P = 0.02]. After treatment, FVC%P values in the OPAT group were greater [98 (IQR = 93.5–107) versus 96 (IQR = 84–107.7), but the difference was not significant (P = 0.082), and there was no statistical difference in the rate of improvement in FVC %P after the treatment between the 2 groups (P = 0.272). As given in Table 2, on admission, the FEF_25–75%P values were also higher in the OPAT group than the hospital group before and after treatment (P < 0.0001 and 0.0012, respectively); both group's FEF_25–75%P values improved similarly (P = 0.98) after treatment. In summary, on admission, lung functions (FEV₁, FVC, and FEF_25–75%P) were lower in the hospital IV group versus OPAT group and remained low at the end of the treatment. There was a trend of better improvement of lung functions in the hospital IV group, but the difference did not reach statistical significance. There was no difference in the time to the first PEx between the 2 treatment groups (184 versus 199 days; P = 0.47). Treatment of 2 PEx episodes in the hospital group was unsuccessful, with failure to return of the discharge FEV₁ to the baseline level and required readmission to the hospital.

Table 2.

Comparison of Pulmonary and Nutrition Variables of Hospital Versus Outpatient Parenteral Antimicrobial Therapy Groups (Median with 1st Quartile and 3rd Quartile) for Categorical Variables

Variable	Hospital IV (N = 242)	OPAT (N = 48)	P
	Median (Q1, Q3)	Median (Q1, Q3)
Weight (admission) (kg)	38.8 (28.17, 49.25)	42.25 (33.7, 52.1)	0.06
Weight (discharge) (kg)	41.2 (29.5, 50.2)	41.35 (33.43, 51.95)	0.46
Weight (discharge–admission)	1.9 (0.9, 3)	0.85 (0.08, 1.47)	0.0001
BMI (admission)	17 (15.5, 19.8)	19.8 (16.88, 21.33)	0.0002
BMI (discharge)	18.2 (16.7, 20.38)	19.05 (17.12, 20.85)	0.3
BMI (discharge–admission)	1 (0.4, 1.4)	0.2 (−0.15, 0.8)	<0.0001
FVC %P (admission)	86 (71.75, 97)	92 (82.5, 99.5)	0.0228
FVC %P (discharge)	96 (84, 107.75)	98 (93.5, 107)	0.0828
FVC %P (discharge–admission)	9 (3, 19)	7.5 (−0.5, 19.5)	0.272
FEF25–75%P (admission)	47.5 (32, 66.25)	66 (54.25, 83.5)	<0.0001
FEF25–75%P (discharge)	65.5 (44, 90.75)	83 (57, 109)	0.0012
FEF25–75%P (discharge–admission)	13 (3, 24)	12.5 (1.75, 27)	0.9846
FEV1%P (admission)	72 (59.75, 84)	80.5 (70.75, 89)	0.001
FEV1%P (discharge)	86 (72, 96.75)	92 (82, 101)	0.003
FEV1%P (discharge)–FEV1 (admission)	12 (4, 20)	8 (2, 22.5)	0.41
Time to first (next) PEx	184 (104, 346)	199 (96, 477)	0.47

n = number of PEx.

BMI, body mass index; IV, intravenous; OPAT, outpatient parenteral antimicrobial therapy; PEx, pulmonary exacerbation.

On admission, the hospital group had a lower median weight [38.8 kg (IQR = 28.17–49.25)] than the OPAT group [42.25 kg (IQR = 33.7–52.1)]; however, this was not statistically significant (P = 0.06). The hospital group gained more weight compared with the home group [1.9 kg (IQR = 0.9–3 kg) versus 0.85 kg (IQR = 0.08–1.47); P = 0.0001]. Similarly, on admission, median BMI values were significantly lower in the hospital group [17 (IQR = 15.5–19.8 (IQR = 16.8–21.3)]. After treatment, median BMI increased more in the hospital group [1 (IQR = 0.4–1.4) versus 0.2 (IQR −0.15 to 0.8); P < 0.0001].

The prevalence of known drug allergy was similar in both groups (P = 0.73). There was no difference in the frequency of new drug allergy in both groups during or after treatment (P = 0.26).

IV antibiotics can cause liver or kidney toxicity or even hemolytic anemia. We tested whether there were elevations in alanine transaminase (ALT), blood urea nitrogen (BUN), and creatinine levels that could be attributed to antimicrobial treatment. There was no difference in aspartate aminotransferase (AST) and ALT levels on admission and after treatment between the 2 groups (P = 0.6). It was notable that BUN elevation occurred more in the hospitalized group (P = 0.005). This could be an indication of better protein intake in the hospital or of kidney toxicity because of administration of more complex and nephrotoxic treatment in the inpatient group. There was no death in either group. There was a single readmission owing to central line infection and another readmission because of displacement of the peripherally inserted central catheter (PICC) line in the OPAT group during a 15-year period.

Discussion

This single CF center retrospective analysis suggests that for a select group of children with CF, OPAT is safe and comparable with the hospital therapy for the treatment of PEx. The majority of children treated in the hospital had Medicaid insurance, suggesting that household income and availability of resources are significant elements to determine the venue of treatment. Admission and discharge median FEV₁%P values were significantly lower in the hospital group suggesting that children with worse lung disease are likely to be treated in the hospital. Hospital treatment improved lung functions more so than OPAT, although this was not statistically significant. Time to next PEx, as a measure of treatment success, was not different between the 2 groups. Hospital treatment was associated with significantly greater weight gain in comparison with the OPAT group.

OPAT for the treatment of CF PEx has gained widespread acceptance because of its advantages over hospitalization such as decreased risk for cross-infection, fewer absences from school or work, less disruption to family's lives, and decreased cost per treatment course.⁴ However, in the home care setting, adherence to the treatment (especially airway clearance) may be suboptimal given the complexity and time burden of the care that in turn may reduce the efficacy of treatment at home.⁹ Although OPAT for the treatment of PEx has been used extensively in adult individuals with CF, most pediatric CF care centers utilize OPAT in select patients and more or less use the similar criteria that our center employs (see the inclusion criteria).

Several adult studies have documented a greater improvement in outcome measures including FEV₁ after hospital treatment in comparison with OPAT for CF PEx.^6,7,18,19 Other studies determined no difference between hospital and OPAT in terms of improvement in FEV₁.^5,8,12 A single prospective study conducted in 1997 with a subject size of 17 (only 31 PEx events) showed no significant difference between OPAT and hospital arms.¹⁴ These contradictory results could be because of the type of study (nonrandomized, retrospective), small sample size, difference in the patient population (mostly adult patients or mixed adult–pediatric patients), insufficient information about adherence to medications and airway clearance techniques at home, and variations in medical practice at different CF centers. So far, 4 Cochrane reviews beginning in 2000, repeatedly emphasized that current evidence related to effectiveness of home IV treatment in CF is restricted to the aforementioned single randomized clinical trial, and suggests that in the short term, home therapy does not harm individuals with CF, reduces social disruptions, and can be cost-effective.^20,21

Most existing publications studied only adults or mixed adult and pediatric patients. There are a few studies on children with CF that compared the outcomes of OPAT with hospital IV therapy. In a retrospective study of 50 children who experienced 143 PEx courses (age 12.7 ± 3.8 years), Nazer et al. reported better outcomes after hospital therapy compared with home therapy.¹⁵ In a European study, 47 children with CF, mean age 13.32 ± 2.9 years, were studied retrospectively. Fifty-four OPATs compared with 77 hospital treatments showed the percent change in FEV₁ and weight gain was comparable in the 2 settings.¹⁶ Through using CF Patient Registry and US CF twin and sibling study dataset, Collaco et al. studied 1,278 courses of antimicrobial therapy in 479 individuals with PEx.¹⁷ Subjects in the hospital group had a greater improvement of lung function immediately after therapy compared with those in the home group; however, the hospital group had a greater decline in lung function with the following exacerbation. In this retrospective study, patients were older (mean age of 18.2 ± 6.5 years for hospital and 22.3 ± 9.4 years for home). Recently, Schechter et al. analyzed PExs recorded in the epidemiologic study of CF by using techniques to control for indication bias to determine whether there is an advantage to inpatient treatment of CF PEx. During the 3-year observation, 2,773 individuals had 4,497 PEx. The median treatment success rate was 74.2% (IQR = 67.9–79.2). Patients treated at centers who relied on inpatient treatment were more likely to achieve successful FEV₁ recovery, indicating a significant advantage to inpatient versus outpatient treatment of CF PEx.²² Our study showed a trend toward greater improvement in lung function in the hospitalized group. It is possible that hospitalized treatment is superior, and because of the limited number of OPAT events and representation of healthier children in the OPAT group, we could not show a statistical difference favoring the hospital IV group. An alternative explanation would be a ceiling effect related to higher starting lung function in the OPAT group.

Maintaining optimal nutrition is one of the major factors contributing to increased predicted survival for patients living with CF.^23,24 Suboptimal nutrition early in life predicts poor lung functions later in life²⁵ and low initial FEV₁%P values.²⁶ A recent study reported that poor nutrition is a risk factor to predict poor response to treatment of PEx in patients with CF.¹⁶ Similar to previous reports,^6,18 our study showed that when PExs are treated in the hospital, patients tend to gain more weight. It is likely that close observation of patients, by addressing their nutritional needs and controlling their blood sugar levels, contributed to better weight gain in the hospital setting. Hospitalization provides a great opportunity to educate patients and families, identify barriers preventing proper care, and address issues related to adherence and suboptimal mental status. Although decreased physical activity in the hospital may be a concern, many centers use physical therapy service run by physiotherapists directed to individual needs.

Although OPAT in CF has been used for several decades, only <20% of children/each year were treated with OPAT at our center. Our center's low prevalence of OPAT may be related to the hospital being the only CF care center in the state of Arkansas. Because of this, some families traveled for 5–6 h to the center for follow-up visits or when complications occurred during the course of treatment. The lack of home care companies in rural areas and limited resources including transportation could be factors in decision making as well. Socioeconomic status (SES) is a strong predictor of clinical outcomes in patients with CF.²⁷ Barriers to access quality care is the key reason for SES-related disparities in CF.^27,28 Previous reports indicated that medically impoverished CF patients were more than 3 times at a greater risk of death, and survivors had significantly worse pulmonary function and growth.^27–29A study showed that while 44.5% of CF patients who were on Medicaid received IV antibiotic treatment for a PEx, only 28.6% of non-Medicaid patients received IV antibiotics suggesting that poor SES are associated with poor lung health.²⁷ The majority of our children who were treated in the hospital carried a health insurance for families with low income and limited resources (Medicaid) and the decision to prescribe OPAT therapy was closely associated with the insurance provider, a marker of SES. Poor SES and limited resources may also be associated with other risk factors such as smoke exposure, poor nutritional status, and lower initial lung functions measured early in life.

Study limitations

Similar to other studies, in this study, children were treated with OPAT when patients and families were deemed eligible for home IV treatment. Consequently, patients with less severe exacerbations and families with better SES constituted the home IV group, resulting in selection bias. However, our center's approach is a common practice used across CF care centers, therefore, the study results may be helpful in showing that OPAT is effective in these select patients. The retrospective nature of our study is another limitation. Because the number of home IV treatments was far less than hospital IV treatments, we used statistical methods to compare the unequal sample sizes. In addition, there was no compliance monitoring at home (especially adherence with airway clearance). However, many centers treat PEx with OPAT only in children with families with a history of good adherence. This study was not designed to investigate the treatment failure rate defined as failure to return to baseline FEV₁ (best FEV₁ value of the past 6–12 months) after treatment.

In conclusion, OPAT for the treatment of PEx was safe and comparable with hospital therapy in a select group of children with CF. However, hospital treatment should be the treatment choice for more sick and vulnerable children and for families with limited resources.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

The authors received no specific funding for this article.