Article Text

1. Thanos Athanasiou1,
2. Vanash Patel1,
3. George Garas1,
4. Hutan Ashrafian1,
5. Kunal Shetty1,
6. Nick Sevdalis2,
7. Pietro Panzarasa3,
8. Ara Darzi1,
9. Sotirios Paroutis4
1. 1Department of Surgery and Cancer, Imperial College London, St. Mary's Hospital, London, UK
2. 2Health Service & Population Research Department, Center for Implementation Science, King's College London, UK
3. 3School of Business and Management, Queen Mary, University of London, London, UK
4. 4Warwick Business School, Coventry, West Midlands, UK
1. Correspondence to Dr George Garas,Department of Surgery and Cancer, Imperial College London, St. Mary's Hospital, London W2 1NY, UK; g.garas{at}imperial.ac.uk

## Abstract

Purpose To determine the association between professors' self-perception of mentoring skills and their academic performance.

Design Two hundred and fifteen professors from Imperial College London, the first Academic Health Science Centre (AHSC) in the UK, were surveyed. The instrument adopted was the Mentorship Skills Self-Assessment Survey. Statement scores were aggregated to provide a score for each shared core, mentor-specific and mentee-specific skill. Univariate and multivariate regression analyses were used to evaluate their relationship with quantitative measures of academic performance (publications, citations and h-index).

Results There were 104 professors that responded (response rate 48%). There were no statistically significant negative correlations between any mentoring statement and any performance measure. In contrast, several mentoring survey items were positively correlated with academic performance. The total survey score for frequency of application of mentoring skills had a statistically significant positive association with number of publications (B=0.012, SE=0.004, p=0.006), as did the frequency of acquiring mentors with number of citations (B=1.572, SE=0.702, p=0.030). Building trust and managing risks had a statistically significant positive association with h-index (B=0.941, SE=0.460, p=0.047 and B=0.613, SE=0.287, p=0.038, respectively).

Conclusions This study supports the view that mentoring is associated with high academic performance. Importantly, it suggests that frequent use of mentoring skills and quality of mentoring have positive effects on academic performance. Formal mentoring programmes should be considered a fundamental part of all AHSCs’ configuration.

• EDUCATION & TRAINING (see Medical Education & Training)
• MEDICAL EDUCATION & TRAINING
• QUALITATIVE RESEARCH
• STATISTICS & RESEARCH METHODS

## Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

## Introduction

Most organisations, from companies and businesses in industry to healthcare institutions, benefit from effective mentoring. In industry, mentoring programmes have been shown to promote employee motivation and commitment,1 enhance organisational performance and lead to business growth.2 Similarly, in healthcare, mentoring has been shown to assist the process of continuous professional development, improve medical training and optimise patient care.3

Mentoring has been shown to benefit the mentees and mentors, and the Academic Health Science Centre (AHSC) in which they work.4 ,5 The benefits to the AHSC are numerous and dynamic including the creation of innovative leaders, the formation of intraorganisation and interorganisation networks, and the cultivation of continuous improvement, which all essentially improve academic performance.4 ,5

For an AHSC to thrive, high academic performance is required to accelerate translational research into improved patient care and provide world-class education. To quantify academic performance of the AHSC faculty, established quantitative measures such as publication output, citation count and the h-index can be considered.3 ,6 Although other measures of academic performance exist, such as research funding, grants, awards and patents, they are more difficult to quantify and have been less well studied to date.3 ,7

A systematic review published in JAMA in 1996 concluded that the actual evidence supporting the fundamental role of mentoring in academic performance is limited.8 The aim of this study was to determine whether researchers' self-perception of quality and frequency of their mentoring skills were associated with their academic performance within an AHSC setting.

## Methods

### Organisational context and study population

Imperial College London is a public university based in the UK. The university specialises in science, engineering, medicine and business, and has been ranked among the top 10 in the world.9 ,10 In 2011/2012 the organisation had a net income of $1.1 billion, of which$480 million was from research grants and contracts, and \$170 million was from government funding which was granted for its distinguished translational biomedical research. Imperial College AHSC was the first AHSC to be established in the UK. Like all AHSCs, it is based on a partnership between a healthcare provider (Imperial College Healthcare NHS Trust) and a University (Imperial College London).11 The Faculty of Medicine is one of Europe's largest medical institutions in terms of size of the workforce, number of students and generation of research income. We used the publicly available university database to generate a list of all the professors in the Faculty of Medicine. We extracted their demographic information including first name, surname, gender, physician status (ie, whether the academic was a physician or not) and the affiliated division or department for each of the participants included in the study.12

### Study design

An online survey was administered to all professors at Imperial College London to assess their mentoring skills. The instrument adopted was the Mentorship Skills Self-Assessment Survey (described in detail below). Statement scores were aggregated to provide a score for each shared core, mentor-specific and mentee-specific skill. Univariate and multivariate regression analyses were used to evaluate their relationship with quantitative measures of academic performance (publications, citations and h-index).

### Survey instrument

The survey instrument administered in this study was used with permission given by CCC/The Mentoring Group, one of four divisions of a not-for-profit corporation: the Coalition of Counselling Centers (Grass Valley, California, USA). The source document for this survey was: Skills for Successful Mentoring: Competencies of Outstanding Mentors and Mentees.13 The survey's research documentation offers evidence for construct validity.

Thirty-six statements from the survey instrument were used to self-assess mentoring skills in the study population (see online supplementary table S1).13 For each statement the participant was asked to indicate the quality (how well—in their personal opinion—they actually performed this skill when it was appropriate) and frequency of application (how often—in their personal opinion—they used the skill, given the opportunity) of their mentoring skills using a Likert scale (see online supplementary table S1).

### Supplemental material

The total score for both quality and frequency of mentoring skills was calculated. Subsequently, the statement scores were aggregated to correspond with each shared core, mentor-specific and mentee-specific skill (see online supplementary table S2).

The survey instrument was based on the shared core skills used by both mentors and mentees, and the unique skills needed by each group. The model postulates that if the shared core, mentor-specific and mentee-specific skills are possessed to a satisfactory quality level, and are used as frequently as necessary, then the chances of having a successful mentoring relationship is enhanced.

### Generation of the publication and citation database

SciVerse Scopus Author Identifier was used as a tool to generate, for each professor, the publication list and the corresponding citations.14 For each professor the following academic performance measures were identified: total number of publications, total number of citations and the h-index. The study time period was from 1 January 2006 until 31 December 2009. The citations that we used were collected in January 2013 in order to allow sufficient time for the most recent publications to be cited. To adjust for individual productivity, for each participant the total lifetime publication record preceding the study time period was identified.

### Statistical analysis

Descriptive statistics were used to present the demographics, performance measures and mentoring variables. Each variable was assessed for normality assumptions using the Kolmogorov–Smirnov test. Fisher's exact test was used to determine demographic differences between responders and non-responders. Bivariate correlation (Spearman's rho for non-normally distributed data) was then used to determine the relationship between independent variables (demographics and mentoring) and dependent variables (academic performance metrics). Data that were non-normally distributed underwent logarithmic (log10) transformation before univariate and multivariate regression analyses were conducted to assess associations between independent and dependent variables. Statistical analysis was performed using IBM SPSS Statistics software V.18.0 and Microsoft Excel 2010. Statistical significance was set at p<0.05.

## Results

A total number of 104 responses were received out of 215 professors that were surveyed, which equated to a 48% response rate. There were no statistically significant demographic differences between the responders and non-responders (p=1.000 for both gender and physician status).

### Correlation analysis

Table 1 demonstrates the bivariate correlations between measures of academic performance and the individual mentoring survey items for both quality and frequency of application of mentoring skills. The quality item ‘I believe I demonstrate or model effective actions that help others learn new skills’ had a significant positive correlation with number of publications (r=0.363, p=0.013) and h-index (r=0.321, p=0.028). Similarly, the frequency item also correlated positively with number of publications (r=0.318, p=0.033) and h-index (r=0.303, p=0.40). The frequency items ‘I believe I know what is important to me’, ‘I believe I recognise my specific weaknesses or growth areas’ and ‘I believe I respect others’ limits and boundaries' had a significant positive association with number of publications (r=0.389, p=0.008; r=0.306, p=0.041; r=0.341, p=0.022, respectively). The frequency item ‘I believe I avoid interrupting others when they are speaking’ correlated positively with number of publications (rho=0.393, p=0.008) and number of citations (rho=0.306, p=0.041). Finally, the frequency item ‘I believe I make optimistic predictions to others about their future successes’ had a significant positive correlation with number of publications (r=0.295, p=0.049), number of citations (r=0.333, p=0.025) and h-index (r=0.325, p=0.028).

Table 1

Bivariate correlation between measures of academic performance and mentoring survey items (quality and frequency) (green: positive correlation; red: negative correlation)

There were no significant negative correlations between any item and any performance measure. Table 2 shows the bivariate correlations between measures of academic performance and mentoring skills (in terms of both quality and frequency of their application). There were no significant positive or negative correlations between academic performance and mentoring skills.

Table 2

Bivariate correlation between measures of academic performance and mentoring skills (quality and frequency) (green: positive correlation; red: negative correlation)

### Univariate analysis

The number of previous publications had a statistically significant positive influence on the number of publications (B=0.651, SE=0.105, p<0.0001), number of citations (B=0.687, SE=0.158, p<0.0001) and h-index (B=0.286, SE=0.068, p<0.0001). There were no statistically significant relationships between the other demographic variables (gender, physician status, scientific field) and performance.

With regards to the univariate regression analyses of the relationship between measures of academic performance and quality of mentoring skills (see online supplementary table S3), only the quality of managing the relationship had a statistically significant positive influence on the number of publications (B=2.224, SE=0.738, p=0.004).

Results of the univariate regression analyses of the relationship between measures of academic performance and frequency of mentoring skills (see online supplementary table S4) suggest that the total survey score for frequency of mentoring skills had a statistically significant positive association with the number of publications (B=0.012, SE=0.004, p=0.006). The following frequency of mentoring skills application items had a statistically significant positive association with the number of publications: listening actively (B=1.728, SE=0.711, p=0.019), building trust (B=2.027, SE=0.883, p=0.027), identifying goals and current reality (B=2.161, SE=0.848, p=0.014), instructing/developing capabilities (B=1.960, SE=0.770, p=0.015), providing corrective feedback (B=1.457, SE=0.581, p=0.016), managing risks (B=1.782, SE=0.717, p=0.002), acquiring mentors (B=1.518, SE=0.538, p=0.007), showing initiative (B=1.209, SE=0.554, p=0.035) and managing the relationship (B=1.873, SE=0.592, p=0.003). In addition, the frequency of acquiring mentors had a positive association with the number of citations (B=1.572, SE=0.702, p=0.030), and building trust and managing risks had a positive association with the h-index (B=0.941, SE=0.460, p=0.047 and B=0.613, SE=0.287, p=0.038, respectively).

### Multivariate analysis

In a multivariate analysis we found that there were no statistically significant relationships between quality or frequency of mentoring skills and measures of academic performance (see online supplementary table S5).

## Discussion

This study has shown that there were no statistically significant negative correlations between any mentoring statement and any academic performance measure. In contrast, several survey items relating to mentoring showed a positive correlation with academic performance. Examples include the perception of an individual's mentoring skills (mentoring skill awareness) and particularly ‘quality of managing the relationship—a mentee-specific skill’, ‘frequency of providing feedback—a mentor-specific skill’ and ‘frequency of building trust—a shared specific skill’ which were shown to have an independent positive effect on an individual's academic performance. The finding that mentoring skills boost academic performance is not novel. The literature supports the view that high quality mentoring is key for one's academic and professional success.15 What is new in this study is that frequency of application of mentoring skills had more positive effects on academic performance compared with quality of those skills.

The fact that frequency of managing the mentoring relationship is important for professorial mentoring perception and academic performance may be an indicator of an individual’s organisational commitment. This is in accordance with previous findings that the amount of time mentors spend with their protégés characterises an effective relationship because it is expected that ‘time spent’ or ‘frequency of meetings’ increases in a longitudinal fashion as the mentoring relationship grows and matures over time.16 Also time and frequency of skill application can be used as surrogate metrics in the evaluation of formal mentoring programmes as characteristics of accountability.16 Finally, the fact that the relationship of frequency of mentoring and performance is positive means first that the two professorial functions of mentoring and research are not contradicting each other in terms of time management, and second that academics in senior roles (eg, professors) can be motivated to participate in formal mentoring schemes.

It has been reported that there must be trust and appreciation between the mentor and mentee for the relationship to yield the best conceivable outcomes.17 Building trust is not a simple process and requires the following: communication, availability, predictability and loyalty, which are part of the mentor-mentee ‘contract’. These skills can be strategically selected as the design focus of organisational mentoring programmes providing senior staff with training on how to establish effective mentor-mentee relationships and how to offer honest, timely and well prepared feedback that takes into account the critical importance of building trust in a stepwise manner. The value of establishing formal mentorship programmes has recently been highlighted in the literature.18 ,19 The benefits of mentoring programmes are not confined to the mentees' academic performance. Mentoring programmes also have a positive impact on the institution's academic performance and even finance.20

In a study carried out by the University of California San Diego, a National Centre for Leadership in Academic Medicine in the USA evaluated a carefully structured academic programme set up in that institution. Interestingly, as part of that programme, reimbursement for both mentors and mentees (5% of their basic salaries) was included to promote participation. It was found that the confidence of mentees in their academic roles was boosted. Moreover, the programme was shown to be cost-effective (despite the reimbursement involved) as it promoted the retention of mentees as faculty when they were ready to make the transition. This cost-effectiveness was a result of human capital retention and the reduced costs for faculty recruitment.20

There are a number of studies explaining what the role of a mentor is.21 Modern mentoring practices, such as e-mentoring, generate more questions than answers on how organisational structures affect mentoring perception. All studies, however, tend to agree on the crucial role that providing feedback and building trust have in the mentor-mentee relationship across diverse organisational settings.

Recently, it has become difficult to initiate or nurture mentoring relationships because of the daily clinical, academic and administrative demands placed on clinical academics. Each of these different areas may require different mentors, which underlines the fact that the mentoring model needs modernisation and adaptation to the current needs related to training, subspecialisation, diversified skills (leadership, team-working and communication), and advances in technology and innovation. Formalising mentoring schemes at organisational, national and international levels will facilitate changes in the mentoring culture and may boost academic performance.3 ,22

### Strengths of the study

The main strength of this study is first the inclusion of an extensive range of determinants of academic performance. Second, the sample size was considerable (ie, over 100 professors from 1 of the top 10 universities in the world), which emphasises the importance of this study for academic healthcare sectors in developed countries. Third, the high response rate for this senior cohort (48%) contributes to the validity of the study. Despite a 48% response rate, which in a senior group like this is considered—in relative terms—extremely strong, it is important to acknowledge that 52% did not respond. This may reflect an element of bias in the results due to differences between the groups, though no statistically significant demographic differences were identified between the responders and non-responders. Finally, it is important to note that the study took place in the healthcare sector, and in particularly within an AHSC where, to our knowledge, no similar analysis has ever been conducted.

### Limitations of the study

The most important limitation of this study is that it is based on self-perceptions of mentoring. This limitation is of particular significance as the literature has shown that physicians have a limited ability to accurately self-assess and, as a result, it has been proposed that the processes currently used to evaluate competence may need to focus more on external assessment.14 ,23 Thus, a more robust study would include a 360° mentor evaluation. Another limitation is the response rate, which by being 48% confines the generalisability of the study findings. In addition, the survey instrument has not been extensively validated in the academic and healthcare sectors. Finally, the outcome measures for performance, such as number of publications, number of citations and the h-index, are not the only metrics that can be used to characterise academic performance. Alternative performance measures for use in future studies of mentoring impact include: financial parameters, grant income, patents and innovation metrics, which are representative of global performance for knowledge management organisations and have implications for business and wider societal impact and organisational resilience.

In conclusion, the findings of this study, based on a survey of more than 100 professors from one of Europe's largest AHSCs, show that mentoring is associated with higher academic performance. Importantly, this study suggests that frequent use of mentoring skills as well as quality of mentoring have positive effects on academic performance. These results suggest the need for establishing mentoring programmes as an integral part of all AHSCs' configuration in order to maximise their faculty's academic performance. Based on parallel examples from industry, the positive effect of mentoring on AHSCs' research output may promote employee motivation and commitment; enhance organisational reputation, net income, performance and innovation, which, in turn, is likely to lead to business growth and improved patient care.

### Main messages

• Mentoring is associated with higher academic performance.

• Frequent use of mentoring skills as well as quality of mentoring has positive effects on academic performance.

• There is a need for establishing mentoring programmes as an integral part of all Academic Health Science Centres' configuration in order to maximise their faculty's academic performance.

• The positive effect of mentoring on Academic Health Science Centres’ research output may promote employee motivation and commitment; enhance organisational reputation, net income, performance and innovation, which, in turn, is likely to lead to business growth and improved patient care.

### Current research questions

• How does mentorship perception of mentors relate to that of their mentees?

• Are there any gender differences in mentoring perception?

• Are there any gender disparities in academic performance?

## Footnotes

• Contributors TA, VP, GG and HA contributed to the conception and design of the study, the analysis and interpretation of data, and the drafting of the article. PP and SP contributed to the conception and design of the study, and the interpretation of data. NS and KS contributed to the conception and design of the study, and to the acquisition of data. AD contributed to the conception and design of the study, and to data analysis and interpretation. All authors contributed to the critical revision of the article for important intellectual content and approved the version to be published.

• Competing interests NS is funded by the National Institute for Health Research via the ‘Collaboration for Leadership in Applied Health Research and Care South London’ at King's College Hospital NHS Foundation Trust, London, UK. The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health.

• Ethics approval Formal ethics approval was not necessary as the study team were not required to access any confidential data, and completion of the electronic survey was taken to indicate consent to participate in the study.

• Provenance and peer review Not commissioned; externally peer reviewed.