Hospital at Home Program for the Treatment of Pediatric Osteoarticular Infections Free

Our hospital, a 180-bed public pediatric facility, has ∼6600 admissions per year. Since the introduction of the HAH program in November 2018, each year, the home-care unit provides treatment to ∼450 patients with acute illnesses and exacerbated chronic diseases. In its first 4 years of operation, 698 patients received intravenous antibiotics, 14.6% with suspected OAIs. It operates as a virtual ward and has the capacity to treat 10 patients at a time.

The HAH staff includes 4 nurses, 1 full-time pediatrician, and 1 part-time pediatrician. Physicians perform routine clinical monitoring of admitted children, with daily home visits made in the morning shift, as needed (at home admission and at least every 3–4 days and also if clinical worsening, fever, new symptoms, drug complications, or antimicrobial therapy changes); when necessary, nurses are available to visit patients in their homes until 9 pm. To guarantee patient care 24/7, nursing staff provide onsite care 12 hours per day and are available for 24-hour telephone contact. Since October 2020, the HAH team has also included a consultant pharmacist and a specialist in pediatric infectious diseases.

The HAH team holds morning meetings to decide which patients should receive face-to-face care at home and which may be monitored by telephone or videoconference based on their clinical status and the procedures they require. Whether or not patients are visited at home on a given day, phone calls with caregivers are scheduled twice per day. Patients requiring in-person evaluation outside of regular daytime hours are referred to the emergency department (ED). The anticipation of adverse events, as well as adequate communication with caregivers, is essential for the success of the project.

Patients with a confirmed or strongly suspected OAI are admitted for intravenous antibiotic therapy (Fig 1). Candidates for HAH are proposed by the attending pediatrician or orthopedic surgeon on the ward, who contacts the HAH pediatrician for assessment. Patients are deemed eligible for the program in accordance with defined criteria (Table 1); subsequently, an individual care plan is devised. Family members undergo face-to-face training from a HAH nurse and receive written and audiovisual materials on topics such as the disease and possible complications, treatments, and procedures (eg, medication infusion systems, venous catheter care). After family members demonstrate an ability to perform the required procedures, the child is transferred to the home environment.

In addition to administering medication to the patient, caregivers are required to record 3 times daily the child’s body temperature, general condition, reports of pain, inflammation at the infection site, and feeding. During home visits, unit staff review these data, perform a clinical examination, obtain samples for any necessary laboratory tests, and check the venous catheters and infusion systems for proper functioning. Catheters are replaced if a malfunction is detected. If the need arises for diagnostic imaging studies or evaluation by an orthopedic surgeon, these are performed in the hospital as inpatients. After improvement is shown using clinical and laboratory tests, intravenous treatment is switched to the oral route and the patient is discharged from the HAH in accordance with the same criteria as in CH (Table 1).

After approval to conduct the study was granted by our institutional review board, a longitudinal retrospective observational study was designed. All patients admitted to the HAH from January 2019 to December 2022 with confirmed OAIs were included. Patients for whom the diagnosis was considered but later ruled out during HAH admission were excluded.

Data were drawn from medical records and an internal database specifically created for patients in the HAH program. We collected information regarding age, sex, referring department, length of conventional hospital and HAH stay, diagnosis, antimicrobial agents used during CH and HAH admission, and vascular access devices employed.

Complications were classified as being related to the device, disease, or drug therapy:

• Vascular access complication refers to those which prompted vascular access removal, such as occlusion, phlebitis, extravasation, and accidental leakage.

• Disease-related complication refers to clinical worsening or failure of antimicrobial treatment.

• Drug-related adverse events (eg, hematologic, hepatic, gastrointestinal, skin reaction).

Hospital readmissions during HAH stay and 30 days after discharge were also recorded.

An economic assessment was performed by the center’s analytical accounting department. Patients diagnosed with OAIs were randomly selected, and the estimated cost of HAH and CH was calculated. A comparison of estimated costs was made between CH and HAH care, excluding surgical procedures. HAH-related cost calculations included nurse and physician salaries (with fixed salaries, regardless of the number of children admitted) and operational costs (eg, transportation, telecommunications, computers, specific infusion pumps, and treatments).

Data concerning the length of CH stay among children with a diagnosis of osteomyelitis from 2017 to 2022 were provided by the hospital billing department. The median length of hospital stay among patients with osteomyelitis from 2017 to 2018 (before HAH opening) was compared with the median length of stay (LOS) from 2019 to 2022 (after HAH opening) to estimate the reduction in time of hospitalization.

A descriptive analysis of all patients included in the study was performed. Quantitative continuous data were expressed as medians and interquartile ranges, whereas categorical variables were reported as frequencies and percentage values. Data were entered into Microsoft Excel 2019 (version 16.25; Microsoft Corporation, Redmond, WA, USA) and analyzed by using SPSS software for Windows, version 25 (IBM Inc., Chicago, IL).

During the study period, 102 patients were admitted to the HAH program with suspected OAIs. These cases represented 49% of all hospital admissions due to SA or osteomyelitis. The presence of an OAI was ruled out in 14 patients during HAH admission on the basis of the findings of advanced imaging tests, and these patients were excluded from this study.

The final sample consisted of 88 children, 53% of whom were male. The median patient age was 4.1 years (interquartile range [IQR], 1.7–10.6). The largest number of patients (n = 58; 66%) were referred from the pediatric ward, 30% (n = 27) from orthopedics, 2 from the ED, and 1 from maxillofacial surgery.

The most common diagnosis was osteomyelitis (n = 50; 57%) followed by SA (n = 26; 29%), myositis (n = 5; 6%), tenosynovitis (n = 3; 3%), discitis (n = 2; 2%), and others (n = 2; 2%).

Twenty-seven patients (31%) underwent some intervention before transfer to HAH, the most frequent being arthrocentesis (n = 14; 54%), arthrotomy (n = 7; 27%), and other surgical drainage procedures (n = 5; 19%). Microbiological culture samples were obtained in 25 (93%) of these patients. A blood culture was taken in 75 (85%) patients before antimicrobial therapy. The diagnosis was confirmed microbiologically in 28 patients (32%), as detailed in Table 2.

The most common antibiotics used in CH and in HAH are shown in Fig 2. In 48% of the patients (n = 42), the antibiotic regimen required modification to allow for HAH admission. A total of 72 patients (82%) received 3 daily doses of intravenous antibiotics, 11 (12%) received 4 to 5 doses per day, whereas the rest (n = 5; 6%) received 1 or 2 daily doses while in the HAH program.

Antibiotics were prepared by the nursing staff and administered by caregivers in 99% of patients. The venous access devices used on admission to HAH, as well as the number and reasons for device replacement, are shown in Table 3. Peripheral intravenous line (PIV) was the most frequently employed device; a single PIV line was used for a median of 4 days (IQR, 3–6).

Blood testing was performed in 90% of cases (n = 79) to monitor disease course or drug side effects during HAH.

Thirteen patients (15%) had an adverse effect of possible drug-related etiology. Five presented with neutropenia (minimum values 890 × 1000/L), whereas 2 patients had mild hepatic transaminitis, and another 4 had a mild self-limited skin reaction. One patient developed diarrhea, although microbiological study results were negative. The remaining patient, an adolescent male treated with cefazolin for osteomyelitis, developed fever and mild renal failure and thus, was readmitted to CH with intravenous fluids and switched to ceftriaxone. Adverse reactions prompted a change of antimicrobial therapy in 2 patients. No additional interventions were required.

Readmission to CH was required in 2 patients (2%) during HAH. The first one was previously described because of renal insufficiency. The second was a preschool-aged girl with SA of the knee who, due to persistent joint inflammation, was admitted to the hospital for surgery and subsequently, returned to HAH to continue antibiotics, with a favorable outcome. One patient with an intraosseous abscess required readmission during the first 30 days after discharge from HAH.

The median LOS in CH of all study patients was 3 days (IQR, 2–5), and the subsequent HAH stay was 7 days (IQR, 4–8.75), avoiding 632 hospital days.

The median hospital stay among patients with osteomyelitis before HAH opening (n = 32) was 8.5 days (IQR, 4.5–12), decreasing to 4 days (IQR, 3–7) after HAH opening (n = 102; P = .005), with a median LOS in HAH for osteomyelitis of 7 days (IQR, 4–9.25).

During HAH admission, a median of 3 home visits (IQR, 2–5) were conducted by the pediatrician, and 5 (IQR, 3–7) were conducted by the nursing staff. A total of 171 unscheduled calls were received, 42% (72/171) during the night shift. Thirteen patients (7.6% of total calls) were referred to the ED; 11 of these were referred because of problems with the device and were able to return to HAH after resolution. The other 2 patients visited the ED because of clinical worsening. The daily estimated cost of CH stays among patients with an OAI was 1044.81 EUR as compared with 342.12 EUR for HAH.

The current study describing our 4-year experience with an HAH program reveals the benefits of home OAI treatment. Our findings reveal that the program provides safe and effective self-administration of multiple doses of intravenous antibiotics, close clinical monitoring, and a reduction in health care estimated costs.

HAH reduces the disruption of hospitalization because it allows personalized care and respects the usual surroundings of children and families. In line with the literature,¹⁹  HAH programs mitigate the negative impact of hospital admission for children and families and decrease anxiety by promoting active parental involvement in the child’s care process.^20,21  In addition, home hospitalization improves communication with families becoming the direct providers of the child’s care. Although the degree of satisfaction was not specifically analyzed in this cohort, based on our experience, families at HAH give highly positive feedback on periodic anonymous surveys.³  Parents report improved rest, feeding, and daily routines, as well as a better work-life balance.

Our HAH program allows almost one-half of patients with OAIs to be treated at home for much of their hospital course. As more experience is gained with the program, earlier HAH admission may be possible, even directly from the ED. Proper patient selection is essential. Our social criteria for HAH admission are designed to identify families with an adequate comprehension level to ensure families properly understand care plans, helping ensure patient safety. In case of language barriers, collaboration with interpreters could be helpful for including such patients in this modality of care because fluent communication is essential to guarantee care continuity.

HAH patients are monitored with similar safety and quality assurances to those in an inpatient ward, enabling the early detection and management of disease-, device-, or medication-related complications.^4,22  Indeed, evidence²³  reveals the beneficial role played by caregivers in detecting medical errors and adverse events in admitted patients.

As described by Bryant,²⁴  the overall duration of medical care may be longer for HAH than for hospital-based treatment, although the daily savings from home admission, 67% in our study, and the greater comfort for patients and families outweigh the disadvantage of lengthier care.

The reduction in hospital stay for OAIs has optimized bed availability in our hospital, with an attendant reduction in estimated costs. However, the introduction of the HAH cannot be the sole reason for this decrease, although there was no change in the hospital protocol for OAIs during the study period. Reductions in LOS may also have a beneficial impact on the risk of health-care-associated infection.²⁵ 

Given the difficulty of analyzing the overall cost during home hospitalization (assessments by physicians other than the HAH specialist, imaging studies, laboratory tests, etc), it is likely that health care costs involved in this model are underestimated. As other HAH cost reviews reveal,²⁶  the intensity of care during inpatient days is not constant because it is higher in the early stages of the disease process, which tend to take place in the hospital setting.

We found a lower readmission rate compared with studies on OPAT,^27,28  and no readmissions were due to vascular access devices. This may be related to the comparatively higher use of PIV lines compared with other series;^4,29  8 of 10 patients were successfully treated with a PIV line as a vascular device, with adequate catheter duration. Potential explanatory factors for this durability are single-person use, a strict maintenance protocol between doses, line use only to deliver antibiotics, and the degree of care taken by parents when performing these procedures on their children. Contrary to other reports,⁴  we have limited the use of midlines and PICC lines, thus lowering the risks and avoiding the need to sedate the child for cannulation. In addition, if replacement was required, PIV line placement could be performed at home. We did not compare the dwell time of PIV lines in HAH versus CH, although doing so would be of interest in future studies.

In 2002, Maraqa et al³⁰  described OPAT as an effective approach to treating OAIs with courses of 42 days for osteomyelitis and 29 days for SA. Our results reveal that HAH remains beneficial in these patients, even with an average stay as short as 7 days. Data from our HAH program reveal a decrease in the length of hospital stay, thereby ensuring that clinical decisions can be made in conditions that resemble those of an inpatient ward.

Adequate antimicrobial selection is important, not only in terms of the required spectrum but also for choosing the optimal dosage and for adequate control of the rate of adverse events. None of the patients in our cohort required a broader antimicrobial spectrum. Multiple daily doses make this possible, for which self-administration is essential in this type of HAH program and feasible for acute pediatric diseases. Carter et al,³¹  reported that self-administration was limited to chronic patients with a recurring OPAT requirement. Changes in elective antibiotics may run contrary to local antimicrobial stewardship programs, a phenomenon Gilchrist³²  referred to as the “stewardship/OPAT dilemma.” In our series, one-half of the antimicrobial agents were switched, always in accordance with pediatric clinical practice guidelines.³³ 

We used cefazolin instead of oxacillin to reduce the daily dose because of its improved profile regarding phlebitis and tolerability. Fernandes et al,²⁷  found oxacillin to have the highest rate of complications in OPAT. As a result, the authors recommend replacing this drug with first-generation cephalosporins, and we support this recommendation.

Sequential oral therapy should not be delayed^18,34,35  because it avoids potential side effects and decreases the overall economic burden of therapy.³⁶  Goldman et al,³⁷  in their retrospective review of 3433 OPAT episodes to analyze the overuse of intravenous therapies, especially for conditions in which highly bioavailable oral alternatives exist, found that 38% of patients required ED visits, 61% of which were OPAT-related. During HAH care, a switch to oral medication could be decided daily.

There are some limitations to this study that should be noted. Because it is a single-center study, the results cannot be extrapolated to other populations or to hospitals with different organizational characteristics. We did not compare the overall outcome of patients who completed all treatment in CH against those who were transferred to the HAH program. The study is descriptive, and it would be interesting to conduct comparative clinical trials with conventional hospital care, which would allow for a clearer demonstration of the advantages of the HAH program over CH. Finally, the pre/post analyses concerning estimated cost and LOS were unadjusted (crude) analyses and, therefore, are unable to account for differences in severity of illness or patient characteristics. This may also represent a patient selection bias that we acknowledge because physicians decided that presumably less sick patients were good fits for HAH while keeping the sicker ones in the hospital. A randomized clinical trial would allow for the avoidance of selection biases.

To conclude, HAH programs for the treatment of OAIs are safe, clinically effective, and cost-efficient options. Regular patient monitoring through daily physician and nurse visits allows for the early detection of complications and avoids ED visits and hospital readmissions. Adequate family training before HAH admission and regular communication are essential to ensure that home admission runs safely.

The authors wish to acknowledge the patients and caregivers of the HAH program, as well as the nursing and medical staff for providing care to these patients.