Outcome of HIV-infected patients transferred to a specialist inpatient unit

Abstract

The British HIV Association (BHIVA) recommends that specialist clinical networks are involved in care of HIV-positive patients admitted to district general hospitals (DGHs) and that transfer to a specialist HIV treatment centre is considered for each patient. We audited our experience of 29 patients transferred to our specialist inpatient unit over a two year period. Fifteen (52%) patients were known to be HIV-infected before admission to the referring hospital. Ten (71%) of 14 patients with newly diagnosed HIV had an opportunistic infection at transfer. At the referring hospital the time taken to diagnose HIV infection ranged from one to 26 days (median = 3.5). Only five patients (17%) were transferred by 72 hours of admission to the referring hospital. The duration of stay at our centre was 1–212 days (median = 15): seven patients (24%) required admission to the intensive care unit. Seven patients died; of these, three had newly diagnosed HIV infection. This audit demonstrates that sick HIV-infected patients transferred to a specialist HIV unit had a poor outcome and lengthy hospital admissions. Our audit supports roll-out of HIV testing to avoid adverse outcomes associated with late diagnosis and development of clinical networks involving specialist HIV treatment centres in order to support provision of HIV care in DGHs.

Keywords

audit HIV AIDS clinical networks inpatients

INTRODUCTION

Provision of specialist inpatient HIV services is variable and is in part dependent on hospital location and local HIV services' caseload. In reality, inner city HIV clinics have a critical mass of staff and patients that enables provision of a specialist inpatient team with experience in management of acutely unwell HIV-infected patients. Smaller services often utilize local general medical expertise, with input from genitourinary medicine and infectious diseases. The British HIV Association (BHIVA) recommends that inpatient care should be coordinated through the development of clinical networks and that for patients admitted to a district general hospital (DGH) under a general medical team (showing no clinical improvement after 48 hours), advice should be sought from a specialist inpatient HIV centre and transfer must be considered.¹

Transfer of patients to specialist units may be delayed due to physician- or patient-determined factors, or due to complex logistics, such as availability of beds. It was our impression that patients transferred to our inpatient unit from other hospitals had a higher mortality than other patients needing hospital admission and who were already accessing care from our outpatient service and so we audited the outcome among this patient group.

METHODS

We audited our experience of patients transferred to the specialist inpatient service at University College London Hospitals (UCLH) from other hospitals between August 2007 and August 2009. The audit ‘bench marks’ were the BHIVA standards: ‘Patients who need inpatient care for opportunistic illnesses, HIV-related tumours or other serious HIV-related disease should ordinarily be admitted to an HIV centre under the care of a consultant qualified to provide HIV inpatient care, or to the relevant tertiary service in liaison with the HIV centre’ and ‘Admission under general acute medical care may be considered for patients with non-opportunistic conditions and otherwise uncomplicated HIV infection. If such patients have not shown a good response to treatment within 24–48 hours of admission, advice must be sought from an inpatient HIV centre, and transfer must be considered’.¹

The hospital electronic clinical data repository at UCLH was used to identify patients transferred from other hospitals. Details of each patient's transfer and subsequent admission to UCLH were obtained by case note review. Information extracted included patients' demographic details, duration of admission at the referring hospital (days), diagnosis made by the referring hospital prior to transfer, whether a patient's HIV status was known prior to admission to the referring hospital and (if newly diagnosed) the time taken to obtain the HIV diagnosis (days), CD4 count and HIV viral load at transfer, duration of admission to UCLH (days) and final diagnosis. The outcome was either died in UCLH, or discharged.

RESULTS

Twenty-nine patients (18 men) were transferred. Their age range was 17–80 years (median = 41); two were >60 years old. The duration of stay at the referring hospital before transfer ranged between one and 33 days (median = 9). Only five patients (17%) were transferred within 72 hours after admission to the referring hospital. Delays in transfer were mainly due logistic reasons, often due to non-availability of beds on the specialist inpatient ward at UCLH.

Fifteen (52%) patients had known HIV infection before admission to the referring hospital. Of these patients, eight were white (6 men), four were black Africans (1 man) two were Asian (both men) and one was an Afro-Caribbean man. In those with known HIV infection CD4 counts ranged from 30 to 1000 cells/μL (median = 280) and HIV viral load ranged from <50 to 360,000 copies/mL (median = 2200). HIV infection had been diagnosed more than three months previously in 13 patients. Four of six patients with CD4 counts <350 cells/μL were receiving antiretroviral therapy.

Fourteen patients had newly diagnosed HIV infection. Eight of these patients were black Africans (5 men), four were white (2 men) and two were Asian (1 man). Among patients with newly diagnosed HIV infection CD4 counts ranged from 0 to 500 cells/μL (median = 20) and HIV viral load was 6000–970,000 copies/mL (median = 115,500). Ten (71%) of the 14 patients with newly diagnosed HIV had an AIDS-defining opportunistic infection at the time of transfer (Table 1). The time taken to make the diagnosis of HIV infection at the referring hospital ranged from 1 to 26 days (median = 3.5). Among patients whose HIV infection was first diagnosed at the referring hospital, the delay in transfer after diagnosis of HIV infection ranged between one and 13 days (median = 7).

Table 1

Details of patients transferred to the specialist inpatient HIV unit at University College London Hospitals (UCLHs)

Patient	Patient's HIV status known prior to admission	Transfer diagnosis	Final diagnosis	Duration of admission at referring hospital (days)	Time to HIV diagnosis (days)	Time from HIV diagnosis to transfer (days)	Duration of admission to UCLH (days)	Outcome
1	No	Cryptococcal meningitis	Cryptococcal meningitis	4	2	2	20	Discharged
2	No	CIDP	CIDP	16	7	9	10	Discharged
3	No	PCP	Bronchiectasis, psychosis	3	2	1	42	Discharged
4	No	Lymphoma or TB	PCP, MAI and HIV encephalopathy (postmortem)	18	5	13	4	Died
5	No	PCP	PCP and multiorgan failure	13	1	12	11	Died
6	No	Cryptococcal meningitis	Cryptococcal meningitis	28	26	2	1	Died
7	No	TB	Disseminated TB	14	5	9	30	Discharged
8	No	Miliary TB	TB and meningitis (S. mitis)	5	3	2	42	Discharged
9	No	Toxoplasmosis	Cerebral toxoplasmosis	10	3	7	39	Discharged
10	No	Brain metastases	Cerebral toxoplasmosis	10	7	3	212	Discharged
11	No	TTP	TTP	1	1	0	9	Discharged
12	No	TB	TB	11	4	7	46	Discharged
13	No	Malaria and pneumonia	Malaria, pneumonia and myeloma	8	5	3	15	Discharged
14	No	PCP	PCP	9	1	8	10	Discharged
15	Yes	Pneumonia	Pneumonia and ITP	11	NA	NA	9	Discharged
16	Yes	Pneumococcal bacteraemia	S. pneumoniae endocarditis	2	NA	NA	48	Discharged
17	Yes	Pneumonia and psychosis	Pneumonia and psychosis	17	NA	NA	6	Discharged
18	Yes	Drug reaction	Drug reaction	5	NA	NA	3	Discharged
19	Yes	Unclear	HIV encephalopathy	1	NA	NA	15	Discharged
20	Yes	Primary CNS lymphoma	Cerebral toxoplasmosis and multi-organ failure	33	NA	NA	21	Died
21	Yes	ITP	ITP and chest infection	14	NA	NA	15	Discharged
22	Yes	Neurosarcoid	TB meningitis	22	NA	NA	150	Discharged
23	Yes	Pancytopenia	Aplastic anaemia	5	NA	NA	56	Died
24	Yes	Burkitt lymphoma	Burkitt lymphoma	5	NA	NA	108	Died
25	Yes	ITP and alcohol dependence	ITP and alcohol dependence	6	NA	NA	7	Discharged
26	Yes	Metabolic acidosis due to chemotherapy	Metabolic acidosis due to chemotherapy	1	NA	NA	5	Discharged
27	Yes	Cerebral toxoplasmosis	Glioblastoma multiforme	13	NA	NA	40	Died
28	Yes	Gastritis	Gastritis	4	NA	NA	7	Discharged
29	Yes	Kaposi's sarcoma	Multicentric Castleman's disease	7	NA	NA	18	Discharged

CIDP = chronic inflammatory demyelinating polyneuropathy; ITP = immune thrombocytopenic purpura; MAI = Mycobacterium avium-intracellulare; NA = not applicable; PCP = Pneumocystis jirovecii pneumonia; TB = tuberculosis; TTP = thrombotic thrombocytopenic purpura

The duration of stay at the referring hospital before transfer was longer among patients with newly diagnosed HIV infection, median (range) =10 (1–28 days) than among those whose HIV infection was already known =6 (1–33 days). The duration of stay at UCLH was 1–212 days (median = 15). Seven patients (24%) required admission to the intensive care unit (ICU). Of these patients, three died (all had newly diagnosed HIV infection). Four other patients who did not require admission to the ICU died.

DISCUSSION

This audit shows that sick HIV-infected patients transferred to a specialist HIV unit had a high mortality rate and often had lengthy hospital admissions. Advanced HIV disease, with associated opportunistic infections represented a significant proportion of this group. Earlier diagnosis of HIV among these patients could have prevented their costly hospital admission.²

Only 17% were transferred to UCLH by 72 hours following admission to the referring hospital. There are several reasons that transfer of HIV-positive patients may be delayed. First, admitting medical teams in a DGH may not be aware of local specialist services or the willingness of these services to receive transfers. Second, logistic factors may delay transfer, for example, a lack of available beds.^3,4 Third, an HIV diagnosis may not be considered by clinicians with a resulting delay in testing and impacting on timely transfer.⁵ This audit identified that delays in transfer were mainly logistic and were frequently due to non-availability of beds on the specialist inpatient ward at UCLH.

Our audit supports roll-out of HIV testing into general medical practice, in order to avoid adverse outcomes associated with late diagnosis.^2,5 Additionally, the audit supports development of clinical networks among HIV treatment centres to facilitate timely transfer of sick patients to specialist HIV-treatment centres, together with closer links between specialist units and general medical teams within DGHs.¹ Increasing awareness of BHIVA standards in DGHs, via specialist networks, is an important step in improving outcomes among this patient group.

A prospective national audit of all transfers from DGHs to specialist HIV-treatment centres would inform decisions about appropriate and timely transfer of vulnerable immunosuppressed patients.

References

British HIV Association. BHIVA Standards for HIV Clinical care 2008. See http://www.bhiva.org/documents/Guidelines/Standards/StandardsHIVClinicalCare.pdf (last checked 13 January 2011)

Fleishman

, Yehia

, Moore

, Gebo

. HIV Research Network: the economic burden of late entry into medical care for patients with HIV infection. Med Care 2010;48:1071–9

British HIV Association. UK National Guidelines for HIV testing 2008. See http://.www.bhiva.org/documents/Guidelines/Testing/GlinesHIVTest08.pdf (last checked 13 January 2011)