Norwegian regular general practitioners’ participation in out-of-hours work

Material and method

The study was conducted in the period May 2006 - February 2007. Information was retrieved for all RGPs (3 915) in the NIA’s RGP overview (10). We excluded 83 nameless recordings and 28 duplicates and sent out the questionnaire to the remaining 3 804 doctors. 47 others were later excluded as their address was not known; they had either ceased to practise, were on leave, sick or had died. As a result, the number of potential respondents totalled 3 757. Two reminders were sent out.

The NIA’s RGP overview contains information on sex, number of persons attributed to each RGP (list size) and number of vacant spaces on the list. In the questionnaire we asked about the doctor’s age, satisfaction with list size, normal number of duty sessions and whether their municipality participated in intermunicipal cooperatives. We then asked the RGPs if they themselves participated in OOH work. RGPs with a reduced number of duties were asked for their percentage of normal OOH duties, and those not participating at all were asked at what age they had stopped. They were asked if they had a regular locum doctor to take on their OOH duties and if they had formal exemption from OOH work.

The RGP was then asked to express their view on 8 statements by marking to what extent they agreed with them on a scale of 5 (completely agree, partly agree, neither agree nor disagree, partly disagree, completely disagree):

Two additional variables were retrieved from Statistics Norway (11): the municipality’s population and its centrality. The centrality is defined as a municipality’s geographical location in relation to a centre where there are important functions (central functions). The centrality is measured on a scale of 0 - 3 where 0 is the least and 3 the most central (12). The municipality’s doctor density was calculated as the number of RGPs per 1 000 inhabitants.

Data are presented as frequency distributions and mean values with 95 % confidence intervals (CIs). Statistically significant differences (p < 0.05) between two groups occur when the point estimate for one group lies outside the CI for the other group. In the analyses, full and reduced duty sessions were put in the same group; likewise complete and partial agreement (and disagreement) with the 8 statements. Also, a logistic regression analysis was done where the outcome variable was whether or not RGPs participate in OOH services.

Results

2 913 RGPs (78 %) responded. Doctors with long lists (>= 1 800) were somewhat underrepresented (69 % replied), and doctors with full lists seemed to respond slightly more often than doctors with open lists (80 % versus 75 %). There were no significant differences in the response rate regarding sex, centrality, population or GP density. The doctors’ mean age was 48 (95 % CI: 48 - 48) years; 45 (44 - 45) years for women and 50 (49 - 50) years for men. The average list size was 1 196 (1 181 - 1 210); 1 106 (1 083 - 1 129) for female RGPs and 1 235 (1 217 - 1 252) for male RGPs. 47 % (45 - 49 %) of the lists were open. 83 % (81 - 84 %) of the doctors were content with their list size; 8 % (7 - 9 %) wanted more patients and 9 % (8 - 11 %) wanted fewer.

64 % (63 - 66 %) replied that their municipality participated in intermunicipal cooperatives and they reported that the normal OOH workload was 3.4 (3.3 - 3.6) duty sessions per month. Municipalities that participated in intermunicipal cooperatives had less OOH workload than those who did not and the difference was most obvious in the smaller municipalities (tab 1).

50 % (48 - 52 %) of the doctors participated fully in OOH work; 15 % (13 - 16 %) participated to some extent; and 35 % (34 - 37 %) did not participate at all. Sex and age distribution is shown in table 2. Those who participated to some extent worked 37 % (34 - 39 %) of the normal OOH workload. The average age for ceasing OOH work completely was 50 (49 - 50) years. 28 % (26 - 29 %) of the doctors were formally exempted from OOH service, and 13 % (11 - 14 %) had a regular locum doctor to take their sessions.

Descriptive analyses show decreasing participation in OOH work with increasing age and list size (and full lists) and increasing municipal centrality and population. There is also greater OOH participation in areas with high RGP density. Doctors who want more patients on their list take on more duty sessions. Doctors who want fewer patients participate less often than those who are content with their list size (tab 3).

Multivariate analyses show that female and older doctors participate less in OOH work and confirm that increasing population and greater centrality are associated with lower OOH participation. In this analysis, list characteristics and RGP density do not seem to matter. Doctors who want fewer patients on their list participate less often in OOH work than doctors who are content with their list size or think they have too few patients (tab 3).

We also did a regression analysis where we only included doctors who either participated fully in OOH work or did not participate at all (doctors participating partly were excluded). This analysis gave the same conclusions.

In a third regression analysis the municipal OOH workload was included as an additional explanatory variable. In municipalities where the OOH workload was at least 3 duty sessions per month, doctors participated more frequently than in those with a lighter workload (odds ratio 1.36; 1.07 - 1.74). Due to insufficient response on this variable, it was left out of the main analysis.

Only 16 % (14 - 17 %) of the OOH doctors were willing to take on extra duty sessions in addition to the ones already allocated to them and 60 % (58 - 62 %) tried to give away many of their sessions. Only 16 % (14 - 18 %) of the OOH doctors under the age of 55 wanted to continue with OOH work after the age of 55, and 10 % (8 - 11 %) after the age of 60.

Most doctors (73 %; 71 - 75 %) felt that OOH work is professionally challenging, but only 28 % (26 - 30 %) saw it as well paid. Doctors in sparsely populated, outlying areas found OOH work more challenging and worse paid than doctors in larger more central districts (tab 4).

75 % (73 - 76 %) of the respondents believed that OOH workload is an important factor for doctor stability in a municipality. Only 6 % (5 - 7 %) wished to move from their municipality. Doctors in remote areas were more likely to want to move than doctors in larger, more central places (tab 4).

Discussion

Strong points in this study are its high response rate and that the material represents the target population reasonably well. Doctors with the longest lists are somewhat underrepresented, but they comprise a relatively small part of the total material. The average list size is the same in this material as among all Norwegian RGPs (13). There was a greater tendency for doctors with full lists to respond than for those with open lists, but the difference was small. We do not know the age of the doctors who did not respond, but the average age in our material is the same as that of all Norwegian RGPs (13). We also do not know the OOH participation rate for the doctors who did not respond, but the participation rate in this study was the same as that previously found in data from the National Insurance Administration (8). We therefore consider the material presented here to be representative for Norwegian RGPs.

Many of the background variables in this study are mutually dependent upon each other. This means that several results, which appear to be significant in the descriptive analyses, loose significance or disappear when corrected for mutual dependence in the multivariate analyses. On the other hand, new significant relations can emerge in multivariate analyses, for example, the correlation between the doctor’s sex and OOH commitment.

Nearly two of three doctors replied that their municipality participated in intermunicipal cooperatives; an organisation form which clearly led to a reduction in OOH duty sessions, particularly in the smallest municipalities. Yet there is nothing to signify that intermunicipal cooperatives causes RGPs to take on more OOH duty sessions. This is similar to the situation in Great Britain. Several studies suggest that larger OOH cooperatives reduce the number of duty sessions for the individual doctor (14 - 17), yet only 10 % of English GPs chose to continue to take on OOH work after they were allowed to opt out (18). Poor salary arrangements have probably also contributed to this development (19).

However, intermunicipal cooperatives have their advantages. They are more robust because more doctors are available to share the duty sessions, and it is easier for older RGPs to withdraw from the on-call duty rota without causing the OOH service to collapse. In addition, intermunicipal cooperatives can provide better working conditions (equipment and staff) and thus offer a better quality of the services (9).

This study confirms that female RGPs and older RGPs participate less in OOH work than others (8). Women opt out of the system sooner than their male colleagues and are more often formally exempted from duty. There is however a marked number of RGPs under the age of 55, both women and men, who do not participate in OOH work even if they are not formally exempted. Many of these clearly have regular locum doctors to take their sessions.

Descriptive analyses suggest that OOH commitment decreases with increasing list sizes and full lists. There is less commitment among doctors who want shorter lists and the reverse for those wanting longer lists. There is also rather more participation in OOH work in areas well served with a high RGP density. These findings suggest that doctors with more than enough to do during normal working hours decline from participation in OOH service. Perhaps there would be more OOH commitment if the RGP density increased and doctor workload decreased (20). The statistical correlations are weakened when corrected for mutual dependence among the explanatory variables. One should be wary of drawing too firm conclusions from these results.

The link between OOH commitment and the municipality’s size and centrality is more robust. There is a clear tendency for lower OOH participation in large, central municipalities than in sparsely populated outlying ones, and this is the case even when one takes into consideration the fact that doctors in larger municipalities are older and have longer lists. One important reason for these differences is presumably that doctors in central municipalities have more opportunities to hand over their duty sessions to other doctors. It is also possible that doctors in outlying municipalities feel a greater responsibility for preventing their small, vulnerable OOH service from collapsing. They experience greater professional challenges in their OOH work than doctors in large municipalities, but are less satisfied with their payment. This is probably one important reason for the high number of doctors that want to move from outlying municipalities.