Hospital admissions from nursing homes: rates and reasons

Abstract

Hospital admissions from nursing homes have not previously been investigated in Norway. During 12 months all hospital admissions (acute and elective) from 32 nursing homes in Bergen were recorded via the Norwegian ambulance register. The principal diagnosis made during the stay, length of stay, and the ward were sourced from the hospital's data register and data were merged. Altogether 1,311 hospital admissions were recorded during the 12 months. Admissions from nursing homes made up 6.1% of the total number of admissions to medical wards, while for surgical wards they made up 3.8%. Infections, fractures, cardiovascular and gastri-related diagnoses represented the most frequent admission diagnoses. Infections accounted for 25.0% of admissions, including 51.0% pneumonias. Of all the admissions, fractures were the cause in 10.2%. Of all fractures, hip fractures represented 71.7. The admission rate increased as the proportion of short-term beds increased, and at nursing homes with short-term beds, admissions increased with increasing physician coverage. Potential reductions in hospitalizations for infections from nursing homes may play a role to reduce pressure on medical departments as may fracture prevention. Solely increasing physician coverage in nursing homes will probably not reduce the number of hospitalizations.

Figure 1

Size and proportion of long-term and short-term beds in all 32 nursing homes in the city of Bergen (250 000 inhabitants), Norway. The red part indicates short-term beds (3-4 weeks stays) and the blue part long-term beds (permanent stays).

Figure 2

Age distribution of patients admitted to hospital from nursing homes in the city of Bergen, Norway, during the period from March 2006 to March 2007 (N = 1311).

Figure 3

Hospitalization rates (per 1000 inhabitants/residents) for total Norway (N = 4.6 millions, age ≥80 years) and among nursing home patients (N = 2300, long- and short-term, average age appr. 85 years), 2007, in Bergen, Norway. (Source: Statistics Norway).

Figure 4

Types of infections on admission from nursing home to hospital in Bergen, Norway, during the period from March 2006 to March 2007 (N = 328).

Figure 5

Types of fractures on admission from nursing home to hospital in Bergen, Norway, during the period from March 2006 to March 2007 (N = 134).

Figure 6

(a) Admissions correlated to physician hours per bed per week (N = 32, R ² = 0.24, P = .0041, y = 0.4396x + 0.4549), grey line with two institutions with high number of short-term beds and extra staffing excluded. (b) Admissions correlated to short-term/long-term beds. Institutions with zero short-term beds excluded from analysis. Grey line with one institution with almost only short-term beds excluded. (N = 15, R ² = 0.55/0.04, P = .0016/0.51, y = 1.6074x + 0.2488 for grey line). Dotted lines are confidence curves for regression lines. The grey circle represents institution 32 (92% short-term beds), and the grey triangle represents institution 31 (43% short-term beds).