Audit of resource use in patients with non‐organic disorders admitted to a UK neurology unit

Patients with non‐organic disorders are commonly seen in neurology outpatient clinics¹ and use a considerable amount of health service resources.² Less is known, however, about the prevalence of these disorders among patients who occupy inpatient neurology beds in the UK.³,⁴

We set out to establish the number of patients with non‐organic neurological disorders admitted to the neurology ward in Oxford, UK, during a 12‐month period and to estimate the hospital resources used by these patients.

Methods

The ward serves a population of about 2.4 million. A relatively small proportion of inpatients are admitted through their general practitioners (GPs). The unit provides a regional and supraregional videotelemetry service. We acknowledge that these factors may limit the generalisability of some of our results.

We studied 693 consecutive neurology ward admissions (563 patients) between August 2001 and August 2002. These admissions were split into two lists on the basis of apparent specificity of the diagnostic coding term (DCT) assigned by the hospital coding office: list 1 (n = 380), DCT organic and unambiguous (eg, Myasthenia gravis); and list 2 (n = 313), DCT non‐specific (eg, epilepsy, unspecified). We predicted that any admissions with a non‐organic diagnosis would be contained within list 2.

After reviewing the medical discharge summary (and subsequent correspondence if necessary), each admission from list 2 was classified (AMP/BM) as being either organic or possibly non‐organic according to our study criteria. The study criteria can be seen online at http://jnnp.bmjjournals.com/supplemental. CB and YH then independently examined all the possible non‐organic cases (n = 58) and classified them as either definitely organic or definitely non‐organic. Admissions were finally classified as non‐organic only if both YH and CB agreed with this diagnosis. For nine cases, either or both of the examiners believed that a non‐organic diagnosis was possible but not definite, and these admissions were therefore classified as uncertain. To determine the accuracy of the DCT, the discharge summaries from a random sample of 80 patients from list 1 were also examined.

For all non‐organic admissions, the following variables were recorded: duration of ward stay, age, sex, referral source, presenting symptoms, investigations used, the information provided to the patient by the neurology team about their diagnosis and whether the patient was subsequently referred for a psychiatry opinion.

For a small number of cases, this latter information was not available (because the relevant outpatient correspondence could not be obtained from another hospital). This is reflected in our analysis by the variable denominator given in the Results section for these outcome measures.

The Mann–Whitney U test was used to determine whether significant differences (p<0.05) existed between non‐organic and organic admissions for age and duration of ward stay.

Results

A case note review of all admissions from list 2 and of a random sample (80/380) from list 1 identified 45 of 313 and 3 of 80 admissions with a non‐organic diagnosis, respectively. The total number of non‐organic ward admissions therefore identified in our sample was 48 of 693 (7%, n = 45, table 1). None of these 48 admissions was given a non‐organic DCT by the hospital coding office. As the DCT error rate on the partially sampled list 1 is 4%, we predict that the true percentage of non‐organic admissions in our total sample of 693 admissions is 9%.

Table 1 Number of ward admissions, patient demographics and duration of ward stay per ward admission in each diagnostic group.

Diagnostic group	Admissions, n	Patients, n	Age (years)	Women, n	Ward stay (days)
Non‐organic	48† (7)	45 (8)	39 (14–74)*	37 (77)*	5 (1–7)
Organic	633† (92)	506 (90)	54 (17–89)	279 (44)	5 (1–147)
Uncertain	9 (1)	9 (2)	30 (17–59)	9 (100)	3 (2–8)
Unknown	3 (<1)	3 (<1)	28 (17–50)	2 (67)	4 (2–4)

Values are n (%) or median (range).

*p<0.001 between non‐organic and organic diagnostic groups.

†Multiple admissions; 3/45 and 90/506 patients with non‐organic and organic diagnoses, respectively, were admitted two or more times during the study period.

In all, 23 of 48 (48%) of the non‐organic admissions were emergencies: 7 of 48 from GPs; 15 of 48 transferred from a district general hospital (14/15 patients first presenting to the Accident and Emergency department); 1 of 48 (non‐epileptic seizure) from the hospital radiology department. Admissions that were transferred to the neurology ward from a district general hospital had a median duration of hospital stay at their referring hospital of 6 days (range 1–21, n = 13).

No marked difference was observed in the duration of ward stay either between non‐organic and organic admissions (table 1) or between non‐organic admissions admitted electively (median 5, range 2–12 days) and non‐organic admissions admitted as an emergency (median 5, range 1–17 days).

In all, 30 of 48 (63%) non‐organic admissions were for possible epileptic seizures (elective admissions, n = 17/25 (68%); emergency admissions, n = 13/23 (57%)). Sensorimotor disturbance in the limbs (including gait disturbance) and movement disorders together accounted for nearly one‐third of non‐organic admissions (29%).

The investigations most often carried out were a standard electroencephalogram recording in 31 of 48 (65%) cases and a period of videotelemetry in 24 of 48 (50%) cases. All 17 elective, and 6 (46%) of emergency non‐organic admissions with a possible history of seizures, underwent a period of videotelemetry during their admission. Videotelemetry seemed to be a useful investigation, capturing a sample of a patient's non‐epileptic attacks in over 80% of cases. Other commonly used investigations included MRI (25%), lumbar puncture (17%) and neurophysiological studies (15%).

Twenty six of 43 (60%) non‐organic admissions were subsequently referred for psychiatric support. The fact that the patient had been told directly by medical staff that his or her symptoms were psychological in origin was recorded only in communications to the GP in just over half—that is, in 25 of 43 (58%) of the admissions. Most of these (18/25 (72%)) patients were subsequently referred to psychiatric services.

Discussion

Consistent with a recent report,⁴ this study has shown that 9% of inpatients on a tertiary referral neurology ward are likely to have a non‐organic disorder. Half of these are admitted as an emergency, often first presenting to an Accident and Emergency department.

These patients are as “costly” as patients with organic disease in terms of the duration of ward stay, and place a major burden on resources. Importantly, as the DCTs assigned by the hospital coding office did not identify any of the non‐organic admissions, the future resource needs for these patients may be underestimated.

A history of possible seizures accounted for most non‐organic admissions. Although videotelemetry was helpful in the investigation, because of resource limitations this was possible only in approximately half of the relevant emergency admissions.

Despite the potential benefits of referral to psychiatric services, 40% of non‐organic admissions were not referred after their diagnosis. Our findings highlight the need for designated psychiatric services for these inpatients.⁵ A more transparent disclosure of the diagnosis directly to the patient by the neurology team may also facilitate the ongoing care of the patient by the GP in the community.

The study criteria are available at http://jnnp.bmjjournals.com/supplemental.