Article Text
PDF
Abstract
Objective Despite the many benefits of paediatric Outpatient Parenteral Antimicrobial Therapy (OPAT) programmes, there are risks associated with delivering inpatient-level care outside of hospital. There is a paucity of evidence defining how best to mitigate these risks. We examined the impact of introducing a dedicated medical team to OPAT, to define the role of increased medical oversight in improving patient outcomes in this cohort.
Design A prospective 24-month pre–post observational cohort study.
Setting The Hospital in the Home (HiTH) programme at Princess Margaret Hospital (PMH) for Children, Western Australia.
Patients All OPAT admissions to HiTH, excluding haematology/oncology patients.
Interventions PMH introduced a dedicated OPAT medical support team in July 2015 to improve adherence to best-practice guidelines for patient monitoring and review.
Main outcome measures Duration of OPAT, adherence to monitoring guidelines, drug-related and line-related adverse events and readmission to hospital.
Results There were a total of 502 OPAT episodes over 24 months, with 407 episodes included in analyses. Following the introduction of the OPAT medical team, adherence to monitoring guidelines improved (OR 4.90, 95% CI 2.48 to 9.66); significantly fewer patients required readmission to hospital (OR 0.45, 95% CI 0.24 to 0.86) and there was a significant reduction in the proportion of patients receiving prolonged (≥7 days) OPAT (OR 0.67, 95% CI 0.45 to 0.99).
Conclusion The introduction of a formal medical team to HiTH demonstrated a positive clinical impact on OPAT patients’ outcomes. These findings support the ongoing utility of medical governance in a nurse-led HiTH service.
- outpatient parenteral antimicrobial therapy (OPAT)
- hospital in the Home
- medical governance
Statistics from Altmetric.com
What is already known on this topic?
There is wide variation in medical governance models for paediatric Hospital in the Home (HiTH) and Outpatient Parenteral Antibiotic Therapy (OPAT) programmes.
Despite the many benefits of OPAT programmes, there are documented risks associated with providing inpatient-level care outside the hospital environment.
There is poor adherence to best-practice guidelines for patient monitoring and review within OPAT programmes.
What this study adds?
The expansion of formal medical governance to paediatric HiTH services improves adherence to best-practice guidelines and improves clinical outcomes for patients receiving OPAT.
The high rates of haematological and biochemical abnormalities detected in patients on OPAT for ≥7 days reinforce the need for close observation and laboratory monitoring.
Line-related complications are frequent in OPAT patients and formal monitoring and documentation procedures are recommended.
Introduction
Paediatric Hospital in the Home (HiTH) programmes, including Outpatient Parenteral Antimicrobial Therapy (OPAT) services, enable children and adolescents to receive conventional inpatient-based medical care at home, facilitating early discharge or avoidance of admission to hospital.1 The advantages of HiTH and OPAT include improved patient and family quality of life, earlier return to school and employment, reduced risk of nosocomial infections, cost savings for health service providers and families and increased hospital bed availability.2–5 Children and families consistently report a preference for receiving care at home.5 6
However, the OPAT ‘virtual ward’, with reduced patient visibility, has potential risks. Clinicians often fail to provide adequate patient oversight7; Hodgson et al identified that 26% of their paediatric OPAT patients received an incorrect dose, route or duration of antibiotics and an inappropriate indication for treatment was cited in 6%.8 Overall, drug-related and line-related complications are reported to occur in 20% to 30% of paediatric OPAT patients, correlating with increasing intravenous therapy durations.4 9–12
Recent international OPAT guidelines advocate for dedicated medical oversight by physicians experienced in infectious diseases (ID), antimicrobial pharmacokinetics and management of complex conditions at home, to ensure appropriate patient selection, suitable antimicrobial therapy, adequate monitoring and comprehensive patient review.1 4 Currently, formal ID and medical stewardship requirements for OPAT programmes are inconsistent among paediatric institutions both nationally and internationally.13 The clinical impact of these expert opinion-based guidelines has therefore not been established.
Princess Margaret Hospital (PMH) for Children in Western Australia is a 240-bed tertiary paediatric hospital with 28 000 inpatient admissions annually. A nurse-led HiTH service including OPAT was established in October 2006, servicing the Perth metropolitan area with a population of approximately 2 million, and encompassing a 40–60 km radius to provide up to three home visits per patient per day. Almost 2% of all PMH inpatients are discharged home with HiTH, with half receiving OPAT. Other indications include sleep studies, post-neonatal intensive care weight/feed support, post-tonsillectomy observations, wound and wet-dressing care, and diabetes mellitus education.
A dedicated medical team was introduced to HiTH in July 2015 to improve OPAT governance, patient review and monitoring rates, in keeping with published UK OPAT guidelines.4 We aimed to evaluate the clinical impact of this medical governance on monitoring rates, adverse events, duration of OPAT admissions and unplanned hospital readmission rates, to create an evidence-based platform to guide future best-practice guidelines for medical input within paediatric OPAT services.
Methods
Study design
We conducted a prospective pre–post, observational cohort study of OPAT admissions at PMH between 1 July 2014 and 30 June 2016. A dedicated OPAT medical team commenced in July 2015; clinical outcomes for children admitted in the 12 months before and after this intervention were compared. In our institution, the patient’s treating consultant decides on whether to refer to OPAT, the choice of antimicrobial therapy and duration of treatment. The ID consultation service provides clinical patient review for ID concerns on formal team request, however it is not a requirement for acceptance onto OPAT.
In the preintervention period, decisions to perform laboratory monitoring or arrange patient review were made ad hoc by the patient’s treating team. In the postintervention period, a dedicated OPAT medical team was introduced to HiTH to provide oversight of OPAT patients’ progress and to facilitate weekly laboratory monitoring (full blood count, urea, electrolytes and creatinine, liver function tests and C reactive protein if appropriate) and weekly clinical review for patients receiving antimicrobials for more than 7 days.4 The OPAT medical team included a paediatric ID consultant and/or a fellow, and a general paediatric consultant or a fellow (total 0.3 full-time equivalent (FTE), with at least one clinician providing input at HiTH nursing handover sessions daily and availability via telephone or email to answer nursing queries Monday to Friday, 08:00–17:00. This team reviewed OPAT antimicrobial prescriptions during the daily round and liaised with treating teams to promote monitoring, advise of laboratory concerns and to provide recommendations regarding antimicrobial therapy, including early conversion to oral therapy or cessation of antimicrobials when appropriate. The general paediatric staff provided oversight and team liaison for non-OPAT patients when required.
Data collection
Baseline demographic information for all patients receiving OPAT during the study period was extracted from the HiTH admission database and cross-referenced with electronic patient results, and included age, gender, admitting specialty, primary diagnosis, duration of OPAT and total length of treatment (time from hospital admission to completion of OPAT). Additional variables included: type and number of antimicrobials used, frequency of antimicrobial administration, type of venous access device (VAD), laboratory results and frequency of monitoring, hospital readmissions and whether a formal ID consultation occurred. Laboratory abnormalities were categorised as neutropenia (absolute neutrophil count ≤1.0×109/L), transaminitis (alanine aminotransferase (ALT) greater than two times the upper limit of normal (ULN),14 ≥80 IU/mL for our reference laboratory) and electrolyte or renal function abnormalities. Preintervention line (peripherally inserted central catheter (PICC) and midline catheter) complications were extracted from the PMH Anaesthesia VAD database. Postintervention line complication data were extracted from the HiTH database and cross-referenced with the VAD database. Haematology/oncology patients were excluded from analyses as the unit has independent antimicrobial protocols and patients frequently have baseline laboratory abnormalities due to underlying disease processes.
Data analysis
Individual patient data were de-identified, entered onto a secure SPSS database and analysed using SPSS V.23.0 (SPSS). Patient demographics and OPAT admission details were characterised using descriptive analyses. Non-parametric numerical data underwent Mann-Whitney U analyses. Categorical variables underwent Pearson’s χ 2 analyses. p values <0.05 were considered statistically significant and 95% CIs and ORs reported.
Ethics
This study was approved by the PMH Governance, Evidence, Knowledge, Outcomes (GEKO, Quality Activities #8237 and #10233) committee and performed in partnership with HiTH and the departments of infectious diseases, general paediatrics and anaesthesia.
Results
There were 502 episodes of OPAT during the 2-year study period (n=262 and n=242 in the preintervention and postintervention periods, respectively), accounting for 4416 hospital bed-days. General paediatrics (208/502, 41.4%), respiratory medicine (116/502, 23.1%) and haematology/oncology (95/502, 18.9%) were the most frequently referring specialties. A reduction in haematology/oncology referrals was observed between the two periods (p<0.001), attributable to cessation of an invasive fungal infection outbreak requiring long-term intravenous antifungal prophylaxis in the preintervention period; all other indications remained stable with non-significant fluctuations. After excluding haematology/oncology patients (n=95), 407 patients were included for further analysis. Among these, exacerbations of cystic fibrosis (CF) or bronchiectasis (17.7%) and skin and soft tissue infections (17.3%) were the primary indications for OPAT, accounting for 1006 (22.8%) and 277 (5.1%) OPAT bed-days, respectively (figure 1). Osteoarticular infections also contributed to a substantial number of bed-days (600 bed-days, 13.6%).
Frequency of admissions for Outpatient Parenteral Antimicrobial Therapy diagnostic groups in the preintervention and postintervention periods and associated bed-days per diagnostic group over entire study period. BSI, blood-stream infection; CF, cystic fibrosis; ENT; ear, nose or throat (otolaryngology) infection; SSTI, skin and soft tissue infection. *p<0.001.
Demographic information for the 407 OPAT episodes is shown in table 1. The overall median age was 5.1 years (IQR 1.8–10.0) and median length of OPAT admission was 5 days (IQR 2–11). Overall, almost half of all patients (180/407, 44.2%) received OPAT for ≥7 days, including 17.7% (72/407) for ≥14 days (figure 2). Nearly 80% (318/407) were prescribed one antimicrobial and 18.4% (75/407) received two antimicrobials concurrently; over three-quarters of these (58/75) were for CF/bronchiectasis treatment. Ceftriaxone, cephazolin and tobramycin were the most frequently prescribed antimicrobial agents (figure 3) and piperacillin/tazobactam with tobramycin (n=20), or ticarcillin/clavulanate with tobramycin (n=18) were the most common combinations. Almost half (43.9%) of OPAT patients received once daily antimicrobial dosing (200/456 documented prescriptions), with 9.0% receiving two times per day, 30.7% three times per day and 16.4% delivered by continuous infusion. Most patients received OPAT via peripheral intravenous cannulae (45.2%, 179/396) or PICC (38.6%, 153/396). Other delivery modalities included surgically inserted central VADs (10.1%), midline catheters (3.5%) and intramuscular injections (2.5%).
Demographics of included Outpatient Parenteral Antimicrobial Therapy episodes with analyses of the preintervention and postintervention study periods (n=407)
Length of Outpatient Parenteral Antibiotic Therapy admission subgroup distribution for the preintervention and postintervention periods (≤3, 4–6, 7–13 and ≥14 days).*p<0.05
Antimicrobials prescribed for patients receiving OPAT. Other: antimicrobials prescribed for less than five patients included acyclovir, amphotericin B, amikacin, cefepime, ciprofloxacin, colistin, daptomycin, metronidazole, moxifloxacin, teicoplanin, tigecycline, vancomycin. pts, patients.
Two-thirds (115/180) of children receiving OPAT for ≥7 days underwent laboratory monitoring, with a median of one monitoring episode per admission (range 1–10). Laboratory abnormalities were detected in 24/115 patients (20.9%), most commonly transaminitis (n=9) and neutropenia (n=9). Other abnormalities (n=6) included: raised creatinine (105 µmol/L with associated haemorrhagic cystitis), thrombocytopaenia (<100×109/L), haemolysis, hyperkalaemia (peak 11.9 mmol/L) and hypokalaemia (2.9 mmol/L). Most abnormalities occurred in patients treated for infective exacerbations of CF/bronchiectasis (36.0%) or pulmonary infections (20.0%). Readmission to hospital was more likely in patients who were monitored (15.2% vs 5.1%, OR 4.23, 95% CI 1.20 to 14.96), however not significantly associated with the detection of laboratory abnormalities. The number of antibiotics prescribed was not associated with a higher propensity for laboratory abnormalities or readmission. One-quarter of patients were formally discussed with ID during their admission in both periods; these patients demonstrated significantly increased laboratory monitoring rates compared with those who were not reviewed (79.7% vs 55.1%, OR 3.32, 95% CI 1.57 to 7.01). Twenty (20/167, 12.0%) line-related complications (figure 4) were identified in patients with PICC (17/153, 11.1%) and midline catheters (3/14, 21.4%), with a greater number of complications observed in the postintervention period (15 vs 5, p=0.019). All midline catheter-related complications involved blocked lines (two preintervention, one postintervention). Five patients with line-related complications required hospital readmission, four ceased antimicrobials early and three converted to oral therapy to complete treatment; eight cases had no outcome documented.
Comparison of line-related complications observed in Outpatient Parenteral Antimicrobial Therapy patients with peripherally inserted central catheters and midline catheters in the preintervention and postintervention periods.
Adherence to monitoring significantly increased in the postintervention period (81.6% vs 47.3%, OR 4.94, 95% CI 2.51 to 9.74). A 55% reduction in hospital readmissions was also observed (OR 0.45, 95% CI 0.24 to 0.86) (table 1). The proportion of OPAT admissions of ≥7 days duration reduced significantly from 49.5% (93/188) to 39.7% (87/219) in the postintervention period (OR 0.67, 95% CI 0.45 to 0.998) (figure 2); the median time from hospital admission to discharge onto OPAT remained stable at 3 days. The median duration of OPAT and total length of treatment were shorter in the postintervention period (6 vs 5 days, p=0.096 and 10.5 vs 8 days, p=0.069), however non-significant on statistical analysis.
Discussion
This analysis of a tertiary paediatric OPAT service following the implementation of a dedicated medical team demonstrated three key results: a 10% reduction in the proportion of OPAT admissions with a duration ≥7 days, improved compliance with laboratory monitoring by greater than 30%, and 50% fewer readmissions to hospital.
It is likely that improved medical oversight contributed to our reduced proportion of patients on longer OPAT courses through encouraging appropriate antibiotic selection and treatment length. As the median length of inpatient stay prior to discharge onto OPAT remained unchanged, this decrease cannot be attributed to prolonged pre-OPAT hospital admissions in the postintervention period. Laboratory monitoring rates significantly increased from below 50% to greater than 80% following introduction of medical oversight, consistent with improvements demonstrated in other studies.9 10 This increased monitoring may have resulted in earlier cessation of antibiotics when laboratory abnormalities were detected by the medical OPAT team. Improved medical oversight was also likely to have contributed to our reduced readmission rate, which fell from 14.9% to 7.3%, which is at the lower end of the broad rate reported in the literature (3.8% to 38%).8–10 12 15
Formal ID consultation rates remained unchanged at about 25% throughout our study and were lower than the ranges reported in other paediatric OPAT studies (30% to 90.5%).7 9 10 16 Nevertheless, we demonstrated improved patient outcomes following the introduction of the OPAT medical team to complement non-compulsory ID consultations. These findings suggest that a dedicated OPAT clinician (either ID or general paediatrics) providing appropriate patient monitoring and review may have a greater impact than formal ID consultation on improving outcomes for OPAT patients, with the additional benefit of avoiding mandatory pre-OPAT ID consultation becoming a deterrent for OPAT referral.
Over 40% of our cohort had PICC or midline catheters in situ during their OPAT admission, with 12% experiencing line-related complications. Our observed increase in line complications in the postintervention period may correlate with greater attention to line monitoring. High line-related complication rates in the paediatric population (10% to 29%) are reported in the literature,9 11 17 18 with one study estimating that 61% of acute healthcare encounters during OPAT were attributable to line-related complications.15 Kovacich et al 19 suggest that one-third of PICC-related adverse events may potentially be prevented by appropriate antimicrobial therapy duration and timely conversion to oral therapies, and recommended ID consultation prior to PICC insertion. Our findings support the importance of monitoring line complications and the necessity of judicious line insertion and limiting dwell time. Additionally, monitoring allows for identification of issues such as increased line faults, and may prompt investigation of product quality/batches or review of line insertion and accessing practices.
This study is among the first to quantify the benefits afforded by medical governance provided to OPAT programmes, even with small dedicated FTE.20 The strengths of this study include the large patient cohort and comparable baseline characteristics between study groups. Our indications for OPAT were similar to reports from other institutions,8 10 15 and the prescribed antibiotics and rates of multiple antibiotic use were similar to other studies,7 8 10 13 15 allowing for institutional comparisons and generalisability of results. Limitations include the inability to obtain the frequency of clinical reviews, individual OPAT or ID recommendations and specific reasons for hospital readmissions. The precise cause of our observed reduction in the proportion of patients on prolonged OPAT and reduced readmission rates could therefore not be established. We were unable to capture laboratory monitoring performed externally, but given this infrequent occurrence, this is unlikely to have impacted significantly on our monitoring estimates. Potential drug-related adverse events were only captured based on haematological and biochemical results available electronically; others such as anaphylaxis, rash or diarrhoea were not recorded. Our reported drug-related toxicity is thus an underestimate. Additionally, our definition of specific blood-related laboratory abnormalities, such as ALT at twice ULN, may not prompt all providers to modify antibiotic therapy, although the authors feel that this level warrants review of the appropriateness of ongoing antimicrobial therapy and consideration of therapy cessation.
The guidelines our institution adopted for patient laboratory monitoring are based on expert recommendations,4 with a firm evidence base not yet established. The practicality, necessity and ideal frequency of some investigations on young patients requires further investigation. While dedicated OPAT medical input improved monitoring rates in our institution, which may subsequently improve the identification of laboratory abnormalities, one could argue that medical oversight should primarily focus on earlier cessation of intravenous antibiotics and timely step-down to oral antibiotics to avoid complications and reduce the necessity for laboratory monitoring. These activities were also prioritised by our OPAT medical team.
Supported by these results, our institution continues to provide an OPAT medical team and has formalised the OPAT referral process. This includes introduction of a prereferral proforma to indicate the antimicrobial treatment plan including oral step-down (if appropriate), frequency of laboratory monitoring and patient reviews. Furthermore, antimicrobial guidelines have been ratified to support short-course therapy for specific clinical diagnoses (eg, cellulitis and periorbital cellulitis), to formalise the early oral step-down process and assist in avoidance of prolonged intravenous courses. Formal cost analysis of these interventions has not been performed, however the projected savings from reduced OPAT duration and readmission rates would be anticipated to facilitate the permanent employment of an ID or general paediatrician as part of OPAT medical support and potentially create cost savings for the health system.8
Conclusion
This study demonstrates that expansion of medical stewardship into a nurse-led HiTH service improves adherence to best-practice guidelines and enhances paediatric OPAT patient outcomes. Increased patient monitoring, reduced hospital readmission rates and reduced duration of OPAT not only confer individual patient benefits, including improved safety through decreased intravenous line and antibiotic exposure, but are also likely to result in health cost savings. Further prospective studies exploring the impact of mandatory pre-OPAT ID consultation alongside dedicated OPAT medical oversight, the impact of patient review frequency and the specific interventions of OPAT medical staff would be beneficial.
Acknowledgments
We would like to acknowledge the PMH HITH nursing team for the management of these patients, the Departments of General Paediatrics and Infectious Diseases for their input into OPAT and its governance, and the families who embrace our service. We would like to thank the Anaesthetic Department for providing access to the venous access device database.
References
Footnotes
Contributors AOM drafted the manuscript, developed the project, contributed to and compiled data, edited the draft and approved the final manuscript. CM, DKY and YPC contributed to and compiled data, reviewed and edited the draft, and approved the final manuscript. JV, ACM and CCB developed the project, edited the draft and approved the final manuscript. ACB conceived and designed the project, reviewed the data, reviewed and edited the draft, and approved the final manuscript.
Competing interests None declared.
Ethics approval Princess Margaret Hospital for Children Governance, Evidence, Knowledge, Outcomes (GEKO, Quality Activity #8237 and #10233).
Provenance and peer review Not commissioned; externally peer reviewed.