Dentists’ Working Patterns, Motivation and Morale - 2018/19 and 2019/20 Methodology

The Dental Working Patterns (DWP) Survey took place during January to March 2020 and all primary care dentists in England, Northern Ireland, Scotland and Wales who had NHS/Health Service activity recorded for either or both 2018/19 and 2019/20 were invited to participate. Following the end of the survey, all returns were validated with the main inclusion criteria to ensure that dentists answered all questions and worked for the full year in question (disregarding any annual leave taken). For the first time, any dentists who recorded fewer than 5 weekly hours of work were also removed from the analysis as they were considered outliers.

Table C1 shows the number of invitations sent to dentists in each country, the total number of returns collected, exclusions and the final sample for both 2018/19 and 2019/20.

Table C1: Total number of exclusions by country, 2018/19 and 2019/20

^{1. As well as the inclusion criteria listed in paragraph C1, dentists in England and Wales who only worked under community/salaried service contractual arrangements during the year were also removed (as discussed in paragraph A3).}

The inclusion criteria mean that the figures in each table of this report are based on a consistent group of dentists for each year. Annex B provides the sample and population sizes for the stratification variables underlying the weighting in each country. Tables C2-C5 show the survey response rates by country, broken down into the variables used to stratify each sample. Note that differences in response rates between the two years of each survey may be partly due to changes in the population.

Table C2: Dental Working Patterns survey response rates by dental type, gender and age, England, 2008/09 to 2019/20

Table C3: Dental Working Patterns survey response rates by gender and age, Northern Ireland, 2008/09 to 2019/20

Table C4: Dental Working Patterns survey response rates by gender and age, Scotland, 2008/09 to 2019/20

Table C5: Dental Working Patterns survey response rates by gender and age, Wales, 2018/19 and 2019/20

^{The response rates are the percentages of dentists invited to participate in the survey who returned a completed survey (whether they worked a full year or not). Please note, a group of dentists in Northern Ireland were unintentionally omitted from those invited to take part in the survey in 2010/11 and 2011/12. These were either new (or returning) dentists or dentists who renewed their registration with BSO between September 2010 and March 2012 and who did not claim a practice allowance between their (re)-registration and the end of the relevant financial year. This meant there was a smaller sample size for the analysis covering those years, as shown in table C3. The impact of not inviting these dentists to complete the survey is covered in the report for those years[1].}

Table C6 and figure C1 show the overall response rates for each country by way of comparison (i.e. the ‘All Dentists’ figures from tables C2-C5).

Table C6: Dental Working Patterns survey response by country, 2008/09 to 2019/20

[1] Dental Working Hours, Northern Ireland, 2010/11 and 2011/12, Experimental Statistics. This report is found at: https://digital.nhs.uk/data-and-information/publications/statistical/dental-working-hours

Figure C1: Dental Working Patterns survey response by country, 2008/09 to 2019/20

Figure C1 highlights that, overall, there has been a gradual drop in response rates for the DWP survey in each country since 2008/09, with two main exceptions:

1. 1. An increase in 2012/13 and 2013/14 for Northern Ireland following lower rates in 2010/11 and 2011/12 due to the administrative error discussed in paragraph C4.

1. 2. A gradual increase in Scotland between the 2014/15 and 2015/16 and 2018/19 and 2019/20 surveys, which means the response rate in Scotland is now the highest it has been since the 2010/11 and 2011/12 survey.

It is important to note that the 2008/09 and 2009/10 survey was the last to use a paper-based questionnaire and that response rates have been lower since then with the introduction of an online collection tool for the 2010/11 and 2011/12 and subsequent surveys. Furthermore, the surveys since 2012/13 and 2013/14 have generally shown an overall reduction in the motivation and morale of dentists in the UK, which may have also contributed to the decline in response rate. However, dentists in Scotland recorded increases in motivation and morale for the last survey, which may also partly explain their increases in response rate for recent years.

As discussed in paragraph A29, the decreasing response rate in England over the last three surveys may have been further exacerbated by sending only email invitations to dentists rather than sending hard copy letters.

To assess whether the composition of the survey responder population is representative of the dental population, the characteristics of the two populations are evaluated in this section.

Please note, unlike the results presented in the main body of the report, the figures in this and the next section (covering tables C7-C12 and figures C2-C3) are based on dentists who worked for either a full or part-year. Without survey data, it is not possible to accurately define this characteristic for non-responders, hence both cohorts of dentists were used to allow more accurate comparisons between the sample and population.

England

Table C7 shows that, overall, the distribution of dentists in 2019/20 was fairly similar for the dental and survey responder population. This can be assessed by comparing the Population (%) and Initial Sample (%) columns to each other, with their difference shown in the final column[1].

There are a few strata with higher levels of over or under representation when compared to others; most notably because younger dentists (<35 years) were less likely to return a survey compared to their older colleagues. However, aside from strata 9 and 13, the strata have percentage point differences of less than 5, suggesting that the population of survey responders is generally representative of the dental population; this helps to mitigate any effects that result from the drop in the response rate. In addition, stratifying the sample to account for the major variables of dental type, gender and age and weighting the average results to the population lessens some of the effects of this variation in the survey response rates.

Table C7: Number and percentage of dentists in each stratum of the dental population and survey responder sample, England, 2019/20

The representativeness of survey data can also be tested by considering average NHS/Health Service dental activity for those who did and did not participate in the survey. As this variable is held for the entire NHS/Health Service dental population in each country the comparison is straightforward, with the results for 2008/09 to 2019/20 shown in the following sections.

England

Table C11: Average units of dental and orthodontic activity by dental type and survey response, England¹, 2008/09 to 2019/20

^{Note: Results are based on dentists who worked for either a full or part-year. ‘Ratio’ column shows the ‘No Survey’ activity figure expressed as a ratio of the ‘Survey’ activity figure for each year. 2017/18 and 2019/20 figures are taken from a mid-year dental workforce dataset. 1. 2008/09 to 2017/18 figures are for England and Wales combined.}

Table C11 shows that those dentists who returned a valid survey had higher levels of NHS dental activity on average. However, comparing the distribution of NHS dental activity in 2019/20 between responders and non-responders shows that the differences between the two groups is not quite as clear-cut as the averages in table C11 suggest as shown in figures C2a and C2b.

Figures C2a and C2b: Distribution of total NHS dental activity by dental type and valid survey response, England, 2019/20

Both charts show quite similar distributions of NHS dental activity for survey responders and non-responders. However, as shown in table C11, the charts demonstrate that those who participate in the survey conduct more NHS dental activity than those who do not respond.

Northern Ireland, Scotland, and Wales

Table C12: Average dental activity by survey response (and per cent difference), Northern Ireland, Scotland and Wales, 2008/09 to 2019/20

^{Note: Results are based on dentists who worked for either a full or part-year. ‘Ratio’ results show the ‘No Survey’ Average Number Courses of Treatment figure expressed as a ratio of the corresponding ‘Survey’ figure for each year. 2017/18 and 2019/20 figures are taken from mid-year dental workforce datasets. Z: 2008/09 to 20018/18 Wales results were combined with England as shown in table C11.}

Table C12 show that those dentists who returned a survey claimed a higher levels of NHS/Health Service dental activity, when compared to those who did not return a survey. However, the differences are generally less for the later surveys when compared to earlier ones. Figures C3a to C3c compares the distribution of NHS/Health Service dental activity between responders and non-responders for 2019/20 in each country.

Figures C3a and C3b: Distribution of NHS/Health Service dental activity by valid survey response, Northern Ireland and Scotland, 2019/20

Figure C3c: Distribution of NHS dental activity by valid survey response, Wales, 2019/20

Figures C3a to C3c indicate that generally those dentists who did not return a survey performed less NHS/Health Service activity in general than those that did return a survey as shown by table C11. Whilst there is a difference in NHS/Health Service activity between the two groups, it is difficult to know how this relates to their actual working patterns.

Once again, it is difficult to predict what effect this will have on the figures. For example, some non-responders may conduct more private work and cover similar total weekly hours compared to those responders showing high levels of NHS activity. This should be borne in mind when looking at the results in the report. Where it has been possible to account for differences in the survey responder population this has been done by stratifying the sample to account for the major variables of dental and contract type, as well as gender and age; this will lessen some of the effects of variable response rates to the survey.