Scientific Papers

Structural characteristics and contractual terms of specialist palliative homecare in Germany | BMC Palliative Care

We first report the descriptive results concerning the structural characteristics of SPHC teams. For each characteristic, we initially assess the respective content from the model contracts. At the end of the Results section, we report the identified structural typology of the team models.


Model contracts

Table 1 shows the characteristics of the model contracts. Saxony and Thuringia have the same model contract [31]. In Hesse, three different health insurance fund groups have different contracts, of which only one is publicly available [25]. Berlin has different contracts for nurses and physicians [21]. The contract for Rhineland Palatinate had been canceled by insurance funds at the time of the assessment [29], and the contract of Saarland is not publicly available, so those contracts were not included in the analysis. The reimbursement scheme was only (at least partially) available for 13 of the 17 contracts (see Table 2).

Table 2 Reimbursement schemes

Most versions of the model contracts date from 2009 and 2010, only Berlin contracts were updated (in 2016).

All the model contracts except for that of Westphalia are based on §132d and §37b of the German Social Code Book V. Westphalia operates with a different palliative homecare model based on general practitioners and palliative consulting teams that offer consultation and, if necessary, coordinate primary and specialist palliative care [33]. In most federal states, SPHC teams enter into individual contracts based on the respective model contracts. In Westphalia, Mecklenburg-Western Pomerania [26], Berlin [21], and North Rhine [28], health insurers and the Regional Association of Statutory Health Insurance Physicians are the main contract partners.

SPHC teams

A total of 196 SPHC teams from all regions provided valid datasets (Table 3). Most SPHC teams were established between 2008 and 2015, after the directive had been passed (see Table 4). Some may have existed previously in a different form but did not state this in our survey.

Table 3 Number of valid datasets and SPHC teams, according to Wegweiser Hospiz- und Palliativversorgung (November 2017)
Table 4 Organisational and medical management of SPHC teams, psychosocial and other professions; other organisational characteristics (n = 196)

Different sources can be used to determine the current number of SPHC teams in Germany [39,40,41], resulting in between 270 and 326 teams (as of November 2018) [42, 43]. None of these sources is completely reliable, so our response rate was based on the data available in the Wegweiser database (270 teams + 13 teams from Westphalia) at the beginning of the assessment. The response rate across the different federal states varied from 32.4% to 100% (see Table 3).

Management of SPHC teams

Model contracts give no specifications about team management.

Nurses were part of the organisational management in more than half of the teams, with one-third being managed by nurses alone (see Table 4). One-quarter were led exclusively by physicians. SPHC physicians held a medical director position in more than half of the teams, while 6.6% of the teams had no medical director.

Team size, team members’ professions and qualifications

Seven of the 17 contracts regulate the minimum number of team members (see Table 1). All contracts demand that staff members must be certified in specialist palliative care or palliative medicine as defined by the directive. In Lower Saxony, cooperating physicians can provide care if they have basic training in palliative medicine [27] (see Table 1).

The teams had a mean number of 30.3 staff members and 10.9 full-time equivalents (FTE), with team sizes varying considerably between 1 and 298 staff members. Many teams did not have detailed knowledge of the proportion of time contributed by their members, especially if physicians from private practices or nurses from mobile nursing services provided SPHC (see Table 5).

Table 5 Full time equivalents (FTE) and number of staff members in total, physicians and nurses (n = 196)

On average, the teams consisted of 10.1 physicians, with the number ranging between 1 and 52, and 19.2 nurses, with a range of 1 to 239 nurses (see Table 5). 37.8% had at least one psychosocial professional as a team member. More than two-thirds of the teams (62.8%) had other employees like coordinators, administrative and office workers, or other therapists specified in free text entries (see Table 4).

In most teams, all the physicians and nurses were certified in palliative care (80.6/79.6%), only a minor proportion had no certified staff members at the time of the study: Two teams had no certified physicians and one team had no certified nurses (see Table 6).

Table 6 Qualifications, activity emphasis, and institutional affiliations of physicians and nurses in SPHC (n = 196)

Activity emphasis of team members

Some contracts determine a minimum weekly working time (or a minimum number of full-time employees in the cases of Hamburg [24] and Saxony/Thuringia [31]) and activity emphasis (North Rhine [28], Saxony/Thuringia) for team members (see Table 1).

In some teams, all the physicians (27.0% of the teams) or nurses (11.2% of the teams) worked less than 50% of their time in SPHC. In 62.8% and 23.0% of the teams, more than half of the physicians and nurses, respectively, worked less than 50% in SPHC (see Table 6).

Institutional affiliation of team members

SPHC teams can employ team members directly, but they can also incorporate physicians and nurses from hospitals or private practices/nursing services.

Model contracts like Bavaria [19] and Saxony/Thuringia determine [31] minimum permanent positions. The model contract of Hamburg [24] determines that at least four FTE nurses must be employed by a nursing service (see Table 1).

Only 40.8% of the teams worked with physicians from only one institutional setting, while 66.1% contained nurses with a single affiliation. Physicians from private practices provided SPHC in 67.3% of the teams, while about 40% of the teams worked with physicians employed by hospitals or directly by the team (Table 6, for more detailed information see Additional file 2: Table 2). Of all the teams, 63.8% employed at least some nurses and 39.8% directly employed all the nurses in the team (see Additional file 2: Table 2).

Organisation of coordination and patient care

The model contracts indicate that on-call services must be provided by nurses and physicians (six contracts), by nurses or physicians (four contracts), by nurses (one contract), or by physicians (three contracts) (see Table 1). Most teams (60.7%) coordinated patient care centrally; those with a decentralised model used, for example, regional satellite teams (see Table 4).

Cooperation with primary palliative care and other professionals

Cooperation with volunteer hospice services is mandatory, as defined in the directive. Other mandatory forms of cooperation are specified in Table 1.

Almost all teams (98%) reported to cooperate with volunteer hospice services. Cooperation with other palliative care providers was also frequently reported (see Fig. 1). Cooperation with non-palliative care professionals or specialists was less frequent. However, formal cooperation contracts were less prevalent (see Fig. 1).

Fig. 1
figure 1

Cooperation and cooperation contracts in specialist palliative homecare teams (n = 196)

Patients under care (in 2016)

For the analysis of patient data, all teams established after 2015 (n = 10) were excluded. Patient numbers should be understood as estimates since some teams could not give exact numbers. On average, the teams cared for 394 patients in 2016, with a considerable range of 40 to 1,712. Almost 80% of those patients died while receiving SPHC (see Table 7). On average, more than 80% of the teams’ patients had an oncological diagnosis. The mean length of care was 33.2 days. On average, the teams reported to drive 21.5 km (up to 60 km) and for 26.3 min (up to 60 min) for a patient visit (see Table 7).

Table 7 Patient care in 2016, n = 186; teams with a founding year ≥ 2016 were excluded

Reimbursement schemes

Twelve out of the 18 reimbursement schemes are publicly available. Table 2 shows the attributes of these schemes [18]. Bavaria has a model contract, but teams negotiate reimbursement individually [19]. Thuringia and Saxonia have model reimbursement schemes, but the fees are individually negotiated [25, 31]. Hesse has two different reimbursement schemes from three model contracts.

All the reimbursement schemes combine different payment models. Case-based lump sums are used in up to nine schemes (depending on the SPHC care level). In some schemes, rates are paid for a certain number of days on the condition that some form of (specified) service is provided (a case-based lump sum per performance day; e.g. a sum is payed for the first 10 days of care, but days are only counted if service is provided). Some schemes are based on weekly (up to three weeks) or daily (up to eight days) rates, but daily rates may only apply when patients are visited. Up to 10 schemes contain fees for services, mainly home visits, but in some cases also phone calls or other services. Rates may be differentiated by the time of service delivery, for instance, during or outside office hours.

Reimbursement schemes in Berlin [21], Brandenburg [22], and Lower Saxony [27] explicitly exclude the funding of mobile nursing services for the provision of treatment care (§37 SGB V) parallel to SPHC for patients in full respectively also partial (Brandenburg) care, meaning that SPHC teams have to cover specialist palliative care nursing services as well as additional treatment nursing services. Basic nursing can be reimbursed parallel to SPHC.

Two schemes allow compensation for travel expenses, and three schemes include fees for GPs and other physicians who cooperate with the palliative homecare team.

Identification of a typology of team models using latent class analysis

Due to our limited sample size, we could only include 14 dichotomised variables in the LCA. For the identification of team models, we chose variables that had minimal missing values, provided sufficient variance and remained meaningful when dichotomised. In total, 186 teams were included in the latent class analysis, 10 teams had to be excluded because of insufficient data. While a three-class model showed the best model fit according to the BIC (see Table 8), we chose the four-class model with a slightly lower fit because of its higher face validity.

Table 8 Latent class analysis—model fit for different class numbers (n = 186)

Table 9 shows the conditional item responses for the four-class model and Fig. 2 shows the variables in the different classes. Class 1 was identified as small independent SPHC teams, characterised by mainly directly employed physicians and nurses. Many of these teams included psychosocial professionals, and their coordination and patient care were centralised. With 77 teams, this was the largest class. Class 2 (n = 49) was identified as large network teams that all had more nurses than the median and worked predominantly with physicians from private practices and nurses from nursing services. Three-quarters reported centralised coordination and decentralised patient care. Small network teams (class 3, n = 42) employed nurses directly but the physicians came from private practices, while some also had decentralised patient care. In class 4 (n = 18), hospital-based teams, both the physicians and nurses were employed by hospitals. These teams showed a high rate of the inclusion of psychosocial professionals, as well as centralised coordination and patient care.

Table 9 Conditional item response probabilities (dichotomised variables)
Fig. 2
figure 2

Team characteristics based on latent classes (n = 186). Percentage of teams (y-axis) with the specific characteristic (x-axis) from the latent classes

Source link